TEACHING LATIN AS a LIVING LANGUAGE: REVIVING ANCIENT, MEDIEVAL, and RENAISSANCE PEDAGOGY for the MODERN CLASSROOM by Daylin

Teaching Latin as a Living Language: Reviving Ancient, Medieval, and Renaissance Pedagogy for the Modern Classroom

Item Type text; Electronic Thesis

Authors Oakes, Daylin L.

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 04/10/2021 04:40:29

Link to Item http://hdl.handle.net/10150/624153

TEACHING LATIN AS A LIVING LANGUAGE: REVIVING ANCIENT, MEDIEVAL, AND RENAISSANCE PEDAGOGY FOR THE MODERN CLASSROOM

Daylin Oakes

A Thesis Submitted to the Faculty of the

DEPARTMENT OF CLASSICS

In Partial Fulfillment of the Requirements

For the Degree of

MASTER OF ARTS

In the Graduate College

THE UNIVERSITY OF ARIZONA

2017

STATEMENT BY AUTHOR

The thesis titled Teaching Latin as a Living Language: Reviving Ancient, Medieval, and Renaissance Pedagogy for the Modern Classroom prepared by Daylin Oakes has been submitted in partial fulfillment of requirements for a master’s degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this thesis are allowable without special permission, provided that an accurate acknowledgement of the source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED: Daylin Oakes

APPROVAL BY THESIS DIRECTOR

This thesis has been approved on the date shown below:

01 May 2017 Dr. Cynthia White Date Professor of Classics

Oakes 2

TABLE OF CONTENTS ABSTRACT ------4 INTRODUCTION ------5 CHAPTER ONE: Grammar-Translation vs. Comprehensible Input ------7 The Current State of Latin Teaching------7 The Natural Approach and Comprehensible Input ------11 CHAPTER TWO: A History of Comprehensible Input in Latin Teaching ------20 Learning Latin as a Foreign Language in the Ancient World (1st – 4th Centuries CE) ------20 The Rise of Grammarians and Commentators in Late Antiquity and the Early Middle Ages (4th- 10th Centuries CE) ------27 The High Middle Ages and Communicative Latin Language Learning (10th-12th Centuries CE) --- 31 The Later Middle Ages and the Rise of the Vernacular (13th-14th Centuries CE) ------38 The Renaissance Humanists vs. the Practicalities (15th-16th Centuries CE) ------42 CHAPTER THREE: Historically Backed Comprehensible Input Methods in the Modern Classroom ------51 A Focus on Sound ------51 Attention to Bilingual Sense Units ------52 Early Introduction to Meter ------54 Grading by Content rather than Grammar ------55 Accepting Learner Language and Non-Classical Vocabulary ------57 A Sample Lesson ------59 CONCLUSION ------62 APPENDIX A: Resources for Living Latin Pedagogy ------68 APPENDIX B: Relevant Latin Textbooks and Their Methods ------69 BIBLIOGRAPHY ------70

Oakes 3

ABSTRACT

This thesis considers the history of Latin pedagogy through the lens of the Comprehensible Input Theory of second language acquisition (SLA) developed by Stephen Krashen in the 1980s. It rejects Grammar-

Translation pedagogy in favor of Living Latin pedagogy, which prioritizes language acquisition over language learning. Evidence of successful Comprehensible Input pedagogy found in many examples of

Latin instruction from history shows the potential to adapt for the modern classroom those historical methods which were oriented towards the acquisition of the Latin language, and these have subsequently been shown to be supported by Krashen’s work.

Oakes 4

INTRODUCTION

The current pedagogical movement variously called Living Latin, Active Latin, Spoken Latin,

Comprehensible Latin, or Comprehensible Input aims to adapt the best practices recommended by second language acquisition research for a Latin classroom.1 When practitioners refer to Latin as a

“living language,” they tend to add qualifiers, such as “living but fixed,” which point to the fact that it is

Classical Latin (literary Latin as it was in approximately 100 BCE-150 CE)2 that is being taught. Even so, their willingness and eagerness to call Latin “living” come from a desire to see Latin as vital and useful, something meaningful in and of itself. The goal of Living Latin pedagogy is to treat Latin as a real language and use it communicatively, rather than teaching it for the grammar, logic, study skills, or SAT help that are often touted (NCSSFL 2003: 147-49). It will be useful, therefore, to analyze a time when

Latin was vital in this way, already no one’s first language, but still the most communicatively powerful second language one could learn.3 Beginning around the first century CE, some teachers of Latin as a second language taught it as a “living but fixed” language.4 In this effort to revive Latin and bring it back into wider communicative use, I will look back at the earliest teachers and those of the periods that followed in an attempt to learn from their pedagogical practices.

The aim of this thesis is not to jettison the modern classroom or modern, second language acquisition research-based pedagogy for Latin language learning. Instead, I describe briefly how Latin is

1 While these names signify slightly different methods/theories, the common goal of each is to guide students to acquire Latin rather than learn it. The definition of acquisition will be discussed below (p. 12) as one of the Natural Approach’s five hypotheses for language learning. 2 There is disagreement about the specific dates. This range encompasses the so-called Golden and Silver Ages: Wheelock 2005: xxxiii-xxxvii. 3 See Too and Livingstone 1998 for a variety of essays analyzing the power-dynamics of classical pedagogy in different historical periods. 4 This terminology is borrowed from Leonhardt (2013). His reasoning for this terminology, discussion of the difference between dead and living languages, and the special place of Latin guided this thesis’ use of the term “living”. While the term is usually defined by the existence of native speakers and organic change in a language’s structures, in the “living but fixed” sense, Latin is considered alive so long as it used for active communication and fixed so long as the basic forms and syntax remain unchanging: Leonhardt 2013: 17-20. Owens also considers how “Living Latin” is potentially misleading (2016: 508).

Oakes 5 taught currently, how the movement for Living Latin differs from those current methods, and the principles of second language acquisition research supporting that movement. I will then look for evidence of teaching according to these current principles in ancient, medieval, and renaissance practices. Where such evidence is found, this thesis will also attempt to examine some of the successes or difficulties of teaching Latin as a living language according to those methods, and it will also consider the potential results of reviving those practices for a modern classroom.

Current formalized research into second language acquisition is not being “conducted using

Latin” (Carlon 2013: 108).5 Teachers of Living Latin see results in their own classes (increased participation, matriculation, and skill), but have no formal, Latin-focused research to rely on as a reason for adopting SLA-supported practices. By analyzing the past, this thesis hopes to point out some historic communicative pedagogy successes that are worth trialing in a modern classroom.

5 The relevance of SLA research to Latin teaching is further problematized because “none has considered reading as the primary goal” (Carlon 2013: 112). Although not all teachers agree on a single, primary goal of teaching Latin, the reading goal has held a special prominence since at least the ACL Classical Investigation of 1924, discussed below on p. 9.

Oakes 6

CHAPTER ONE: Grammar-Translation vs. Comprehensible Input

The Current State of Latin Teaching

A major shift in the history of the Latin language occurred between 1750-1850 CE, when Latin lost its practical use as a communicative world language, but remained a major component of education

(Leonhardt 2013: 245). At the time, Latin study was considered part of the foundation of a humanist education. Latin was valued for its ability to provide access to classical texts, ancient political parallels, and as a mark of class.6 Latin study was part of a general trend towards classicizing and historicizing. This revived love for the classical period was also intertwined with a new love for science, and “philology” was redefined as a scientific understanding of language.7 This turn of events was extraordinarily significant for the teaching of Latin and the presence of the language in society. This was when “Latin came to be viewed as an essential tool for training students in grammar and logic; it was also used to separate the good students from the bad” (Leonhardt 2013: 271).

“Latin teaching nearly always included a massive dose of grammar,” but the rise of German philology in the late nineteenth century led, even in the secondary schools, to “grammatical hypertrophy” (Wacquet 2001: 146-47).8 Between 1820 and 1880, classical philologists began to create a systematic understanding of the syntax of Latin and to write numerous grammars (Leonhardt 2013:

273). The earliest extant prescriptive or descriptive grammars were written in late antiquity, when Latin first began to decline as a spoken language. The philological grammars of nineteenth century Europe were different in that they made attempts to “derive inductively a system that would clarify the internal

6 Languages and dialects of languages are too often problematically tied to perceptions of class and even moral judgements of their speakers. Latin’s own place within this tradition is well elaborated in Too and Livingston’s collection of essays on this topic (1998). Waquet also includes a chapter on the class associations of Latin: 2001: 207-29. For a more specific history of the connection between Latin and class during the nineteenth and twentieth centuries see Campbell’s 1968 article. 7 Philology in the broad sense of studying language had been practiced alongside Latin at least as far back as the classical period, but the new German sense of philology was dedicated to discovering the logical underpinnings of the language and delving deep into questions about syntax (Leonhardt 2013: 272). 8 Hypertrophy is usually used to describe enlargement of an organ as a result of enlargement of the cells of an organ. In this context, as each cell of grammar study enlarged, grammar study increased overall.

Oakes 7 coherence and logic of the Latin language” (Leonhardt 2013: 273). Unlike the prescriptive or descriptive grammars which provided basic instructions and examples, the speculative grammars created detailed rules and procedures by analyzing every element of a Latin sentence. Leonhardt argues that in many ways this was a critical point for Latin. Now, Latin was more a language for specialists and “for many who felt forced to learn the language in school, the exercise was about as pleasurable as going to the dentist” (2013: 278).

Wheelock’s Latin, revised and reprinted numerous times, has long been a standard text for the first year and follows the Grammar-Translation methodology. Frederic Wheelock began writing the text in 1950. He aimed to make the early lessons of Latin more interesting for students who now might study only one introductory year of Latin in college rather than come to college prepared to read Cicero and

Caesar. His innovation was to introduce “Sententiae Antiquae” and Etymology sections to make the daily grammar plod more interesting to students. A standard lesson from Wheelock introduces new forms

(e.g., Ch. 14: “I-stems; Ablatives of Means, Accompaniment, and Manner”), explains the use of those forms in English, and offers a vocabulary list of approximately twenty new words. The text then offers artificial “Practice and Review” and “Sententiae Antiquae” sections. The former contains sentences written specifically to include the new grammar and the later slightly edited sentences and passages from ancient authors. The chapter ends with an Etymology section and some English commentary on why Latina est Gaudium – et Utilis! (‘ Latin is a Joy – and Useful!’)

The Grammar-Translation method is very often criticized for its tedium. It is a slog of memorization, methodic parsing of text, finding main verbs and subjects, considering the sequence of tenses, and applying rigid formulas for different types of conditions. Grammar-Translation is primarily

Oakes 8 about decoding, not reading. Even at the beginning of this method’s rise,9 teachers objected to this shift away from interacting with all aspects of the Latin language.

In the first half of the twentieth century, W.H.D. Rouse attempted to promote use of the Direct

Method10 for teaching classical languages, but the method never gained traction, partly due to backlash from proponents of Grammar-Translation.11 In 1924, a report was published by the American Classical

League following an investigation into the position of the Classics in American schools. The section of the report on the Direct Method concluded:

It is undoubtedly true that remarkable results have often been obtained by experienced teachers working with the Direct Method under favorable conditions. Nevertheless, we do not recommend that this method be adopted for general use … (ACL 1924: 234).

The central complaints against the method were the lack of training available, its inapplicability to large classes, the slow progress and inability to transfer students to other methods, the ineffectiveness of untrained teachers, and its limited aims (ACL 1924: 234-35). Fundamental to the investigation was

“determining the validity of ultimate objectives” (ACL 1924: 33). The study included no oral or productive language objectives, but focused on “ability to read new Latin after the study of the language in school or college has ceased” (38).12 Further, the study recommended oral and written techniques as

“a means to secure a thorough mastery of the inflectional forms assigned to the work of each semester”

9 Various dates exist for the slow development of this method. It has its initial beginnings in the sixteenth century, developed slowly as Latin became more of an intellectual study, and by the nineteenth century became the “standard way of studying foreign languages in schools” (Richards and Rogers 2014: 4-8). 10 The Direct Method involved the use of Latin as the sole classroom language, removing the focus on translation. The Natural Method gave rise to the Direct Method. The Direct Method in turn led to other methods. All the methods in this lineage attempt to have students learn language the same way that they learn their native language, with little to no use of the first language. Typically, listening and speaking skills are taught before reading and writing skills; grammar, later or not at all. A few Direct Method offshoots are reading methods (only the reading skill is taught), the Audiolingual Method (memorized dialogues), and a school of communicative approaches (ability to communicate is prioritized above all else). Direct Methods are contrasted with “grammar- based” approaches, of which Grammar-Translation is the most prominent. See Richards and Rogers 2014 Approaches and Methods in Language Teaching for further discussion. 11 See Stray 2011: 5-7. 12 Other objectives that were studied related to Latin’s ability to improve English abilities, disciplinary objectives related to students’ “habits of thinking,” and cultural objectives: ACL 1924 33-35.

Oakes 9

(ACL 1924: 231). As Latin teachers saw their remit reduced to teaching students to read, a continuous scholarly tradition of reaction and re-reaction occurred. The report itself defined reading as the “ability to read Latin as Latin, that is, to get hold of the sense in the Latin order without translation” (ACL 1924:

189). Carr (1928) reacted to the study only a few years later by asking, “Shall We Teach Our Pupils to

Read Latin?”, and attempted to explain why the Grammar-Translation method was so ineffective if true reading was the objective.

It is my firm belief that as long as the teacher habitually uses translation into English as the chief if not the only way to test the pupil’s understanding of a given sentence or paragraph his pupils will use English words and English word-order as the chief if not the only means of getting the meaning of that sentence or paragraph. (Carr 1928: 506).

Carr accepted the ability to “read Latin as Latin” as the ultimate goal, but argued that the only way to truly achieve that goal was with a method for “teaching pronunciation, vocabulary, forms, and syntax functionally, and one that will lead the pupil to realize that language is something used for conveying thought” (1928: 510).

In 1935, Mitchell reacted to Carr’s argument for a lesser focus on technical grammar and translation in the classroom.

If these so-called fine distinctions in case, tense, and mood are valueless for reading Latin, why do they exist in Latin? ... They are an integral part of the Latin language. They mean something. How is the pupil to translate the text containing them unless he knows them, and how shall he know them except he be taught them? Every one of these delicate constructions has a meaning and a force which must be brought out in the English into which they are to be rendered. (Mitchell 1935: 195)

In 1970, the debate was still ongoing over the true meaning of “Reading Latin as Latin.” That year,

Cracas cited Carr’s rejection of both transposition and translation as he defined teaching objectives that could be measured in students. As “a ‘reader’ deals with the language directly in its own terms … If a person can proceed in this manner and then answer basic comprehension questions based on the story line, we say that this person is reading Latin as Latin” (Cracas 1970: 85).

Oakes 10

Also in the 1970s, the first Latin textbook was published which followed the Inductive Reading

Method, the Cambridge Latin Course. Other reading texts employing this method, e.g., Latin via Ovid,

Ecce Romani, and Oxford Latin Course have been published since.13 They all teach the grammar inductively by presenting examples of grammatical rules in an initial reading to guide the student to extrapolate the rules. The readings are followed by explanations of the grammar and decoding or translation exercises. Like Grammar-Translation, they follow a strict progression of introducing grammar features and new vocabulary.

In contrast to Grammar-Translation and the Inductive Reading Method, an alternate tradition of teaching grew using Hans Ørberg’s Natural Method textbook, Lingua Latina, published in 1955. The

Natural Method was a second language teaching methodology developed by L. Sauveur at his Boston- based language school in the latter half of the nineteenth century, and which was formalized by 1901

(Richards and Rogers 2014: 11). Ørberg’s adaption of the Natural Method was rooted in the idea that students should be able to understand the language as it was presented to them. There are no English explanations or translations in the text. Ørberg subsequently supplemented and updated his text and although it has gained a devoted following of Latin teachers and popularity at the college level,14

Grammar-Translation and Inductive Reading methods are still in wide use. In the modern foreign languages, however, communicative approaches like the Natural Method are generally employed. Such approaches are expected to be based on a specific theoretical conception of how language is learned.

The Natural Approach and Comprehensible Input

First and second language acquisition research (relatively new in the 1950s) improved its own methodologies and formalized generally agreed upon precepts or observable facts about how a first

13 See Appendix B for a full listing of Latin textbooks cited and their methods of instruction. There is potential to use these textbooks in Reading Method or Comprehensible Input classrooms. That they involve a greater amount of input is an improvement on Grammar-Translation, but the way the books are structured, i.e., around grammar topics and translation rather than communicative content, is problematic: Macdonald 2011: 4. 14 See Appendix A: Scholia or Carter 2011: 21.

Oakes 11 language is acquired. Researchers began to consider seriously how closely second language acquisition followed these same processes. These theories became the basis for an updated version of the Natural

Method. The foundational text is The Natural Approach: Language Acquisition in the Classroom, by

Stephen D. Krashen and Tracy D. Terrell (1983). The theory for language acquisition laid out by Krashen and Terrell has been modified over the years. Different versions of the method now exist, but the central hypotheses remain. These hypotheses were mainly supported by research into how children acquire English as a second language. That research and the development of the hypotheses were performed by numerous researchers, but Krashen selected five hypotheses which he felt had to be considered when developing a method of second language teaching.15

The first hypothesis of Krashen’s16 theory of second language acquisition is related to the central complaint and discussion behind the movements for Living Latin discussed earlier: there is a significant difference between “knowing a language” (e.g. having passed a class) and actually knowing a language for communicative purposes (e.g. spontaneously using the language in conversation). As most Latin teachers are concerned with students’ ability to interact with classical texts, this boils down to complaints about whether students are actually able to read Latin or only able to decode it. Krashen’s first hypothesis is called the acquisition-learning distinction (Krashen and Terrell 1983: 18-21, 26-27 and

Krashen 1985b: 1). This hypothesis draws a line between knowledge about a language and ability in a

15 “Numerous articles as well as well as full-length books since that time have been devoted to rejecting both the theory and research Krashen used in support of the Natural Approach … as well as its practical applications:” Richards and Rogers 2014: 273. Despite its lack of widespread acceptance in the modern languages (partly because the type of teaching it advocates is not easily aligned to age-/level-based standards), the majority of proponents of Living Latin base their pedagogies on Krashen’s hypotheses at some level and they remain widely discussed. See Masciantonio 1988: 53; Macdonald 2011:1-2; Carlon 2013: 107-109; and Patrick 2015: 109-111. 16 Krashen was the primary force behind the Second Language Acquisition theory in the Natural Approach, Terrell the force behind practical classroom pedagogy. Moving forward, the theory had more influence than Terrell’s methodology, and when people speak of the text now, they privilege Krashen’s contributions. Krashen also has written numerous texts and articles over the years that modify and further explicate the implications of his theories, while Terrell’s scholarship has tended to be focused in other directions.

Oakes 12 language. Knowledge is learned consciously, ability is acquired subconsciously,17 and learning does not lead to acquisition.18 In essence, learning the rules of Latin does not lead a student to acquire ability in

Latin. Learning and acquisition are separate processes in the brain and require separate kinds of teaching. It is clear that accepting this distinction leads to a rejection of the traditional Grammar-

Translation methodology. Acquisition is defined as a “subconscious process for developing ability in language” (Krashen 1985b: 100). This distinction is the first criterion for defining a practice as successful in this thesis’ study of historical Latin pedagogy that follows. If a practice aims at acquisition, then subsequent hypotheses are analyzed. If a method is rooted in learning, it is rejected.

Krashen likewise rejects the Grammar-Translation method is his natural order hypothesis

(Krashen and Terrell 1983: 20, 28-30 and Krashen 1985b: 1), which states that grammatical structures are acquired (not learned) in a predictable order. Observations of children acquiring English as a second language have shown that “certain structures tend to be acquired early and to be acquired late”

(Krashen 1983: 28). This would seem to support language instruction that follows a specific order in teaching grammar structures, but the implications of the hypothesis are more nuanced. Further research reveals that “there is evidence that it [the order] is independent of the order in which rules are taught in language classes” (Krashen 1985b: 1). Students will acquire a structure when they are ready to acquire that structure. Sometimes they will leap ahead; other times they may remain at one level for a long time. The knowledge that language acquisition has a natural order19 has important instructional

17 The terms conscious and subconscious are outdated. More current terminology refers to “explicit” and “implicit” knowledge or the difference between “meta-knowledge” and “knowledge”. In the language acquisition sphere this distinction is exemplified by the fact that students for whom English is their acquired first language may or may not be able to define subject-verb agreement, but their subjects and verbs still agree. This thesis, rooted as it is in understanding and applying Krashen’s research, will retain his original terminology. 18 Krashen’s theory for what does lead to language acquisition is the input hypothesis discussed below, p. 15. 19 The natural order is not known in its entirety for any language. In general, however, it is clear that across several languages and language learning circumstances, we can point to some structures which are mastered earlier than others. For English, Krashen provides the example average order of acquisition: 1. -ING (progressive), Plural, Copula (to be). 2. Auxiliary (progressive), Article (a, the). 3. Irregular Past 4. Regular past, 3rd Singular (-s), Possessive (-s) (Krashen and Terrell 1983: 29). Notice that the irregular past is generally acquired before the

Oakes 13 implications. Attempting to force a late structure to be acquired early (perhaps through error-correction or rule teaching) could prove ineffective or even detrimental.

Krashen’s monitor hypothesis (Krashen and Terrell 1983: 30-32 and Krashen 1985b: 1) deals with the proper place of error-correction and knowledge of grammar rules when producing language.

Ability comes from subconscious knowledge; editing, however, uses conscious learning about language.

The monitor could also be called the editor. An utterance must first be produced through the language that has been acquired. Sometimes the acquired system will also self-correct, but the learned language monitor works only to make corrections (Krashen and Terrell 1983: 31). Often when students are being asked to produce something beyond their acquired capabilities, they will fall back on their first language

(L1) + the monitor. There is a place for the monitor in a language classroom, but it requires first that the student “has to have enough time,” “has to be thinking about correctness,” and “has to know the rule”

(Krashen and Terrell 1983: 30). The monitor will do a better job with rules that are both easy to learn and easy to apply than with complex or advanced grammar which is easy to forget or requires

“elaborate or complex movements of permutation. An easy rule in this sense is the English third person singular which only requires the attachment of a morpheme -s to the end of certain verbs” (Krashen and

Terrell 1983: 31). Even so, the students must be concerned about being correct before they engage the monitor. This does not mean that teachers should encourage a high amount of concern about correctness in their students. Such concern leads to learning, not acquisition, as the next hypothesis explains.

The emotional or attitudinal factors that affect a student’s performance in the classroom are called the affective filter (Krashen and Terrell 1983: 21, 37-39 and Krashen 1985b: 1). The affective filter acts as a barrier between the neurological processes that are working to acquire language and the

regular past. The natural order does not necessarily align with traditional ideas of the simplicity or difficulty of a structure.

Oakes 14 classroom activities meant to facilitate language acquisition. Unmotivated, self-conscious, or stressed students (those with a high affective filter) may participate in class but still not benefit from the practice.

Motivated students who are engaged in the activity (those with a low affective filter) will be most open to language acquisition. While some of these factors are outside the teachers’ or students’ control,

Krashen asserts that “the filter is down when the acquirer is not concerned with the possibility of failure in language acquisition and when he considers himself to be part of the group speaking the target language” (Krashen 1985: 3-4).

Krashen’s final hypothesis and the one that is most central is called the input hypothesis

(Krashen and Terrell 1983: 32-35 and Krashen 1985b). Krashen argues that language is acquired in only one way, “by understanding input that is a little beyond our current level of (acquired) competence”

(Krashen and Terrell 1983: 32). The idea that comprehensible input is the central requirement for language acquisition is Krashen’s main contribution to SLA pedagogy and the source of the core terminology of the Living Latin movement: Comprehensible Input (C.I.). Krashen’s hypothesis argues that

“listening comprehension and reading are of primary importance in the language program, and that ability to speak (or write) fluently in a second language will come on its own with time” (Krashen and

Terrell 1983: 32). He further argues that acquisition according to the natural order comes from understanding input at the current level of acquisition (i) and one step beyond it, also known as i +1.

Because the focus is on understanding that input, input that is i+2 or 3 will not be comprehensible and not lead to acquisition, but rather the L1 + monitor learning that is discussed above.20 Krashen studies caretaker language21 to conclude that i+1 is not difficult to achieve, provided the focus is on comprehension. If those supplying the input are constantly checking for comprehension, the i+1 level where comprehension is achievable for the student will naturally emerge. The focus must be on active

20 See p. 14. 21 E.g., the language used by a child’s caretaker to communicate with the child: Krashen 1985b: 4.

Oakes 15 communication22 and use of context and extra-linguistic information to make sure that communication is occurring; as Krashen states, “we use meaning to acquire language” (Krashen and Terrell 1983: 32).

A two-way interaction allows for better acquisition in that it allows for a “negotiation of meaning, and thus gives the acquirer optimal data to work with, i.e. comprehensible input” (Krashen

1985b: 34). As forcing production causes conscious learning rather than acquisition, other forms of interaction must be employed early on. A silent period will exist for students as they begin to acquire input and make meaning of the new language. The Natural Approach calls for teachers to accept this silent period and use other ways of checking comprehension, such as Total Physical Response, yes/no responses, and extensive use of gestures/visuals (Krashen and Terrell 1983: 76). This period can be short or long for individual students, but forms the first stage of language acquisition and language teaching.

Production will occur subsequently as learners experiment with this “learner language,” that is, the language the learner has acquired. The “learner language” may have norms different from the target language, but it is still a functioning language. Initially only single words or short phrases will be produced. Production will lag behind comprehension.

Following basic language teaching founded in the hypotheses discussed above, Krashen recommends sheltered teaching. “The sheltered class is a subject-matter class made comprehensible for the second-language student,” but it must be about real, meaningful subject matter to ensure that focus is on the message, not the medium (Krashen 1985b: 71). After sheltered-teaching, a student may be ready to mainstream in some content areas (i.e. take courses with students who are native speakers or perform real world language tasks), but not in all. As it takes time for a student to acquire new subject vocabulary or discourse, a survey course, for example, will be difficult as students constantly struggle to

22 N. 10 places communicative approaches within the overall category of the Direct Method. Comprehensible Input is a communicative approach because it is focused on acquiring language through communication above all else.

Oakes 16 adjust to new authors. Survey courses should be held off for last-stage learners. In the early stages of acquisition, an important feature is narrow input and with great frequency.

In the charts below Krashen lays out this plan for foreign language study for both high school/university and elementary levels of foreign language instruction:

(Krashen 1985b: 78)

Oakes 17

(Krashen 1985b: 87)

Both of Krashen’s plans aim for students to be able to participate fully in content-based learning by using their second language after periods of general learning and sheltered learning. In his high school program, Krashen allows for more extensive grammar study than in his elementary program. He places grammar study in the context of monitor use (i.e. knowledge of grammar is useful for editing) or as subject matter. The place of grammar study in Comprehensible Input is one of the most controversial elements of Krashen’s hypotheses.

While Krashen initially argued that input is enough for the acquisition of grammar, he acknowledges that not all input is processed all the time. Students will ignore that which is not necessary for creating meaning. As vocabulary is critical for understanding, it can sometimes be sufficient for understanding; in this case grammar will not be processed. Later research by VanPatten23 on how students process input and turn it into output led him to develop a grammar teaching technique called Processing Instruction, which complements Krashen’s method and is often taught side-by-side with it.24 According to this technique, input must be structured in such a way that understanding the grammatical form is essential.

23 See VanPatten 2003: 25-41. 24 See Carlon 2013: 109-10.

Oakes 18

Thus, teaching must be structured with some level of focus on form. Which forms are important and the depth to which they should be taught remains debated.25 Other research on Krashen’s theories and more nuanced Comprehensible Input teaching leads most language teachers to combine methodologies in their classroom. Many different modern teaching practices have been created under the

Comprehensible Input heading.26

The following is a summary of Krashen’s hypotheses. These will be applied in Chapter Two of this thesis, which studies historical teaching practices through the lens of Comprehensible Input.

• Learning is not acquisition and acquisition is the goal. • Acquisition happens through meaningful Comprehensible Input. • Teachers must constantly check that input is comprehensible and i+1 is achieved. • Student stress, disinterest, or pressure to produce activates the affective filter and impedes acquisition. • The place of grammar instruction is peripheral and primarily used to aid the monitor.

25 See Ruebel 1996, “The Ablative as Adverb: Practical Linguistics and Practical Pedagogy” (63) and Mahoney 2004, “The Forms You Really Need to Know” for two potential solutions (3). 26 See Appendix A for various resources created by practitioners of Living Latin pedagogy. Practitioners are not completely in agreement with Krashen’s theories, cf. n. 16.

Oakes 19

PART TWO: A History of Comprehensible Input in Latin Teaching

Learning Latin as a Foreign Language in the Ancient World (1st-4th Centuries CE)

By the first century CE, the Latin language had, in some sense, become a fixed language. The rapid language shifts which had characterized early Latin mostly ceased. Particularly as a literary language, where permanence and idealization of earlier writers reigned, Latin became a language with a standard grammar. In the second half of the first century BCE, Varro wrote his grammatical works. These and others of the same period served to standardize the basic patterns of inflections (Leonhardt 2013: 60).

As the Roman Empire grew, the desire for non-native speakers to learn Latin also grew, and the tradition of teaching Latin as a foreign language with fixed grammar began. Aulus Gellius (c. 125 – after 180 AD) provides the example of a learned man from Spain:

Adulescens e terra Asia de equestri loco, laetae indolis moribusque et fortuna bene ornatus et ad rem musicam facili ingenio ac lubenti, cenam dabat amicis ac magistris sub urbe in rusculo celebrandae lucis annuae, quam principem sibi vitae habuerat. Venerat tum nobiscum ad eandem cenam Antonius Iulianus rhetor, docendis publice iuvenibus magister, Hispano ore florentisque homo facundiae et rerum litterarumque veterum peritus. (Aulus Gellius, Noctes Atticae 19.9.1-2

A young man from the land of Asia, of the equestrian rank, pleasing in character and morals and equipped well by fortune and with an easy talent and disposition for the musical art, was giving a dinner for his friends and teachers in a little rustic place close to the city for the celebration of the annual day, on which was his first day of life. There had come then with us to the same dinner the rhetorician Antonius Iulianus, a public teacher for young men needing education, a man with a Hispanic pronunciation, but of flowery eloquence and expert in our early literature.27

This passage from Gellius is evidence that Latin was being learned as a foreign language (Adams 2003:

16), but the best evidence we have for how Latin was taught in the first through fourth centuries CE is through the existence of language learning texts that were used by Greek speakers in Egypt. These texts were preserved due either to the survivability of papyri in the desert or through their bilingual nature, which caused them to be preserved by medieval Latin speakers hoping to learn Greek. Eleanor Dickey’s

27 All translations are my own unless stated otherwise.

Oakes 20

2016 work, Teaching Latin the Ancient Way: Latin Textbooks from the Ancient World, combines selections from papyri28 with selections from the Hermeneumata Pseudodositheana. The

Hermeneumata is a collection of bilingual learning materials (glossaries, a colloquium, a phrasebook, and some reading texts) which were grouped into a textbook. Though the textbook now exists in nine different versions, they are called together the Hermeneumata Pseudodositheana.29 As some versions of the textbook have the grammar of Dositheus attached, the texts were initially thought to have been written by Dositheus. Dositheus’ grammar (written in the fourth century CE) was added to the original bilingual textbooks because, despite the tradition in the ancient world for grammars to be written in the language they described, Dositheus composed portions of his grammar for presentation in both languages.

The place of grammar and thus of monitor learning in Latin teaching during this period is partly discoverable through Dositheus’ fourth-century CE bilingual grammar. Dositheus seems to have been working from a monolingual Latin source; he provides the bilingual format for those parts which seemed to him most important for his beginning students. In his section on cases, he spends little time on the cases that are also used in ancient Greek, but instead concentrates on the Latin ablative (Dickey 2016:

88). He provides the names of the other cases, a simple definition of the ablative case, and many translated examples of its use. This indicates a desire to provide the minimal amount of grammar necessary for comprehension and communication.

28 For full dating and locations of the papyri used by Dickey, see Dickey 2016: 176-182. I relied entirely on her summation and reproduction of the texts. 29 Eleanor Dickey is the primary scholar and editor of the texts contained in the Hermeneumata. She uses evidence from the texts to determine the dating for the various portions of the collection. The earliest text is a bilingual schoolbook, written in the West in the first century CE or earlier. The schoolbook moved to the East and a preface and glossaries were attached. A phrasebook written for Greek speakers in the second century was attached to some versions. The schoolbook became mostly fixed by the fourth century, but additions continued to be made from the fifth century onwards, mostly in the western tradition of the text. Dickey isolates parts of the colloquia which were probably produced in the West for learners of Greek from those written in the East for students of Latin, primarily based on the topics described in them. Latin schoolchildren learned Greek, while Greek adults learned Latin: Dickey 2012 I: 32-43.

Oakes 21

The bilingual nature of these texts promoted their preservation, but weakened their efficacy for

Direct Method teaching, including the Natural Approach based on Krashen’s hypotheses. While providing a translation is one method to make text comprehensible, a translation also reduces the amount of input that a student is receiving. Current approaches usually aim to make the input comprehensible by simultaneously using more active techniques; if these were used by teachers in the first to fourth centuries CE, evidence of such activity has not been preserved. What is preserved, however, is not inconsistent with Krashen’s hypotheses in several important ways.

The first thing to note is that the goal of the language teaching was language acquisition, not language learning. Translations were not the goal of learning, but instead provided as a means of checking comprehension and making the input comprehensible. The texts themselves are also meaningful for communicative exchanges. Many of the texts are colloquia, texts written to be transcribed, memorized, and spoken aloud by students. They are in the first person or in dialogue form and on topics relevant to the students’ interests. Most Latin language learners in the East were primarily interested in the language in order to interact with the Roman army or legal system; some of the others wished to make a good impression on the aristocracy when visiting Rome (Dickey 2016: 2-3).

Making topics meaningful to students is a key element of Krashen’s hypotheses. He argues that acquisition is only possible when a student is engaged with the language and that content that is not meaningful has a negative effect on the affective filter. Some topics described in the colloquia include a trip to the bank and the clothes market, a visit to a sick friend, criminal trials, a lawsuit, resolving a dispute, recovering debt, reading a letter from a friend, having a guest for lunch, visiting the baths, a dinner party, being scolded for drunkenness, and going to bed.30 The setting for all of these is in a

Roman context, in order to teach culture alongside language. For example, correct behavior in a Roman

30 Dickey extracts colloquia on each of these topics from the various Hermeneumata manuscripts (which she had previously edited in her 2012 work) for her 2016 work: Dickey 2016: 11-27.

Oakes 22 bath would be important for a Greek hoping to move up in the world, and the passage about the baths features a character going through the entire process. These texts are highly artificial, but written with a view towards comprehension rather than grammar acquisition.

The following passage is from the start of a colloquium on a court case from the Colloquia

Monacensia-Einsidlensia (4a-e).31

Dickey 2012 Vol. I: 110-111

This passage shows some nonstandard grammar features, particularly the second declension vocatives

(4a line 5 Gaie, b line 2 Lucie), but is written with a view towards being as comprehensible as possible for the Greek speaker. The text is presented in columns, with one column of Greek chunked into word or

31 The Colloquia Monacensia-Einsidlensia is one of the nine versions of the Hermeneumata colloquia Dickey edited for her 2012 work. Translations are hers.

Oakes 23 sense units and another with a similar Latin sense unit. The author of the text wrote the colloquium to be presented in this fashion and was careful to avoid structures which could not be understood with an equivalent Greek expression. The author does not avoid, however, using unfamiliar Latin grammar structures, provided they are equivalent to the Greek sense unit. The focus remains on meaning over structure. Unlike Grammar-Translation which attempts to build a vocabulary with careful consideration for introducing successive readings, the concern of these authors was for vocabulary that would be useful for the student. In this case, several offices of the Roman courts are described. The repetitive question structure is also a key feature to help increase comprehension, as repetition both builds understanding and supplies a quantity of input.

As outlined above, following the schoolbook colloquium, some versions of the Hermeneumata contain sections of a phrasebook written originally in the East. The sections are arranged according to topic and likely meant to be memorized and then deployed in the right circumstance. The following comes from a portion containing phrases for excusing oneself, found in the Colloquium Harleianum

(15e-g).32 Section g resembles a substitution drill practiced in a modern Comprehensible Input classroom to help students focus on forms in which the same introduction or closing to a sentence is used; as in the modern drills, each student is prompted to complete the sentence in a different way.33

32 The Colloquium Harleianum is another of the nine versions of the Hermeneumata. Dickey 2012 II: 27-28. 33 See Lindzey 2015: 80 for a sample lesson.

Oakes 24

Following the colloquia and phrasebooks, columnar bilingual versions of summaries of the Iliad and

Aesop’s fables are included in the Hermeneumata, and also the Judgements of Hadrian34 (Dickey 2016:

58-75). The first two are traditional learning texts; the third may have been included to appeal to future lawyers, who sought legal and practical texts. Tying instruction in this way to the interests of one’s students is a key element of teaching according to Krashen’s hypotheses, as content is a primary motivator for language learners. There is also a difficult technical treatise on manumission, which would have been too complex for beginners, but relevant to the interests of those students learning Latin for legal purposes. This is also evidence of an attempt to grade35 texts for various levels of readers.

Apart from the Hermeneumata, evidence of Latin learning in this period is preserved on papyri, some of which are provided in Dickey’s 2016 work. Some papyri include evidence of the need for Greek students to spend time learning the Latin alphabet (Dickey 2016: 119-20), but there is evidence that some students, who only wanted conversational Latin, relied on transliterated texts of colloquia, phrasebooks, and conjugations, and were not concerned with being able to read Latin (Dickey 2016:

121-27). The Hermeneumata texts were probably all written with bilingual columnar presentation in mind, but such presentations also exist on papyri for a few original Latin texts, such as Cicero’s In

Catilinam I and portions of the Aeneid (Dickey 2016: 144-48). The translations are less helpful for these texts because the columnar Greek equivalents are more awkward, as the two versions were not written

34 The Judgements of Hadrian were a series of vignettes which show Hadrian giving his judgements in person. They may or may not be historically accurate. 35 “Grading” is creating a program of increasing difficulty, which allows students to slowly work up to more advanced texts or concepts.

Oakes 25 at the same time with this presentation in mind. The line by line presentation, however, serves as an aid to make the texts more comprehensible.

Up to the early Middle Ages, Latin and Greek texts were generally presented without punctuation or word breaks, and words sometimes flowed across lines. It takes a good deal of familiarity with a language to be able to read such texts. The columnar presentation, however, allows an untrained eye to much more easily distinguish word breaks and sense breaks. Indeed, this presentation of ancient texts (“syntactic-chunking” or simply “chunking”) is becoming a fairly popular method now for teachers seeking to make texts less intimidating and more comprehensible for their students. Such a presentation, however, is usually only used with texts that would be too difficult otherwise and, as with some of the ancient ones found on papyri some difficulties arise from the fact that these were not written with such a presentation in mind.

Finally, some papyri provide evidence of advanced students attempting to read complete texts

(Dickey 2016: 81). These include notes on elements important for comprehension, an unfamiliar vocabulary word for example, or the addition of a few important macrons (particularly on e and o, which the Greeks distinguish in their alphabet by the use of two separate letters, one for the long vowel [η and

ω, respectively] and one for the short [ε and ο]). Dickey assumes these texts were to be read aloud and translated (2016: 81), but it seems possible that the notes were also meant simply to aid the readers’ comprehension. Either way, the distinctions indicate the importance of sound to ancient readers.

It is almost impossible to know what level of language production ancient Latin teachers asked of their students. Absence of evidence is not evidence of absence, but a main element of the Grammar-

Translation method is the composition of simple disconnected sentences,36 and no examples of such activity have been discovered (Dickey 2016: 116). A papyrus of the third or fourth century CE, however, has translations of Greek fables into Latin in a fashion that could represent efforts at extended prose

36 See Appendix B, Wheelock and above p. 8.

Oakes 26 composition (Dickey 2016: 116-18). Other than that, there are some examples in the schoolbook portion of the Hermeneumata of oral recitation and parsing exercises (non-spontaneous or limited forms of production). These portions of the text, however, were probably produced in the West and do not provide reliable evidence for the learning of Latin as a second language.

What evidence exists through the Hermeneumata and papyri is that of a tradition of language teaching which embraced several key features of Comprehensible Input teaching, most notably its desire to provide meaningful input for its students. The presentation of materials also seems designed to aid comprehension: the texts are pre-chunked according to sense units and accompanied by simple first language translations. Some evidence of grading exists in the texts, e.g. avoiding constructions that are not easily glossed. The focus is not on the presentation of grammatical features, but rather on the presentation of useful vocabulary and cultural information. A system of increasing difficulty relies upon unfamiliar content: first daily activities, then conversations, arguments, stories on familiar topics, legal matters, simpler presentation of original Latin texts, and finally the Latin texts themselves.

The Rise of Grammarians and Commentators in Late Antiquity and the Early Middle Ages (4th-10th

Centuries CE)

Through the sixth century CE, Latin was the language of the army and the legal system of the

Roman empire. Papyri evidence in the East gathered by Dickey shows a continuation of the tradition of bilingual Latin teaching already discussed; the latest text Dickey supplies is possibly from the seventh century CE (Dickey 2016:179-82). In the West, however, evidence for the teaching of Latin as a second language remains scarce, although evidence for the teaching of Latin as a literary language is evidenced by the types of grammars produced. The grammars which had been in use before the fourth century aimed to describe the spoken language. As the spoken language grew more distinct from classical literay

Oakes 27

Latin,37 grammars were now needed which taught the literary form of the language. (Leonhardt 2013:

96-102)

Donatus wrote a grammar in the fourth century to fill this gap for his students. Donatus’ grammar foregrounded understanding the letters, parts of speech, and morphology tables. Before students could begin to approach a theoretical understanding of grammar (i.e., a study of how the language worked) they needed to learn the fixed morphology used by the authors of the classical texts they were reading.

Such grammars do not fit comfortably within the category of second language textbooks; rather the grammarians were attempting to “fix” the errors that had crept into the vernacular.38 A base knowledge of the language was assumed, but a “corrected” classical form of the language needed to be taught. The grammars provided the morphology of this formal version, but in order to learn syntax, word usages, and vocabulary, young students still practiced by reading classical texts. The texts were the input, but

Donatus’ morphology along with first language knowledge was what made it comprehensible.

Donatus’ grammar also included information on pronunciation, e.g. explanations of the i which is sometimes pronounced as a j, or what to do with that anomalous letter h (is it an aspirant?) (Donatus,

Ars maior 1.2). The first section of Donatus’ Ars minor39 covers nouns, including what they are, what their properties are, and how they are declined. He covers the property of gender with this series of questions and answers:

genera nominum quot sunt? quattuor. quae? masculinum, ut hic magister; femininum, ut haec musa; neutrum, ut hoc scamnum; commune, ut hic et haec sacerdos; est praeterea trium generum, quod omne dicitur, ut hic et haec et hoc felix; est epicoenon, id est promiscuum, ut passer, aquila. numeri nominum quot sunt? duo. qui? singularis, ut hic magister; pluralis, ut hi magistri.

37 Clackson and Harrocks 2007: 227, 272-92 for a description of the language changes that occurred from the early empire to its fall in the West (476 CE). 38 The desire to “fix the language” is especially evident in the third section of Donatus’ Ars maior, which includes the subsections: barbarismus, solecismus, cetera vitia, metaplasmus, schemata, tropi: Copeland and Sluiter 2009: 83 39 The Ars minor, circulated as a basic grammar text, was composed for beginners using excerpts from the Ars maior, both texts were written around 350 CE: Copeland and Sluiter 2009: 82-103. Latin text from http://www.thelatinlibrary.com/don.html.

Oakes 28

How many genders are there of nouns? Four. What are they? Masculine, as hic magister (this teacher); feminine, as haec musa [‘this muse’]; neuter, as hoc scamnum [‘this stool’]; common, as hic and haec sacerdos [‘this masculine and this feminine priest’]; there are also nouns of three genders, which are called “everything” as hic and haec and hoc felix [‘this masculine happy one and this feminine happy one and this neuter happy one’]; there is the epicene, that is ‘indiscriminate,’ as passer [‘sparrow’], aquila [‘eagle’]. How many numbers of nouns are there? Two. What are they? Singular, as hic magister [‘this teacher’]; plural, as hi magistri [‘these teachers’]. Following the section on nouns, the Ars minor included declination paradigms. Donatus produced a descriptive grammar, which provided examples rather than rules or logic to justify his teachings. His grammar was intended to provide the classical version of Latin to speakers of vernacular Latin.

One of the most famous Latinists of the fourth century was Augustine of Hippo, who grew up in

North Africa in a thoroughly Romanized colony. Latin was his first language; his only second language,

Greek, was very poor (Brown 1996). In discussing his early education, he reflects upon learning the two languages:

Cur ergo graecam etiam grammaticam oderam talia cantantem? nam et Homerus peritus texere tales fabellas, et dulcissime vanus est, et mihi tamen amarus erat puero. credo etiam graecis pueris Vergilius ita sit, cum eum sic discere coguntur ut ego illum videlicet difficultas, difficultas omnino ediscendae linguae peregrinae, quasi felle aspergebat omnes suavitates graecas fabulosarum narrationum. nulla enim verba illa noveram, et saevis terroribus ac poenis, ut nossem, instabatur mihi vehementer. nam et latina aliquando infans utique nulla noveram, et tamen advertendo didici sine ullo metu atque cruciatu, inter etiam blandimenta nutricum et ioca arridentium et laetitias alludentium. didici vero illa sine poenali onere urgentium, cum me urgeret cor meum ad parienda concepta sua, id quod non esset, nisi aliqua verba didicissem nona docentibus, sed a loquentibus, in quorum et ego auribus parturiebam quidquid sentiebam. hinc satis elucet maiorem habere vim ad discenda ista liberam curiositatem quam meticulosam necessitatem. (Augustine, Confessions 1.14)

Why therefore did I even hate Greek grammar, singing such things? For, Homer also is skilled in weaving such tales, and is most sweetly empty, and nevertheless to me as a boy he was bitter. I believe for Greek boys Virgil also is like this, when they are forced to learn him, as I learned him: for difficultly, the difficulty of learning a foreign language at all was sprinkling as if with bile all the Greek charms of the fabulous stories. For I knew none of those words, and with savage terrors and punishments, so that I might know them, I was threatened vehemently. On the other hand, even at one time as an infant certainly I knew no Latin, and nevertheless I learned by noticing without any fear and torture, and even from the sweet cajoling of my nurses and the jokes of those laughing and the happiness of my playmates. Truly I learned those things without the punishing

Oakes 29

burden of those urging me, since my own heart urged me towards producing its own ideas, which it would not, unless I had learned some words not from those teaching, but from those speaking, into whose ears I was also bringing forth whatever I was feeling. From this it is clear enough that free curiosity has greater power for learning those things than dreadful necessity.

Augustine ventures that perhaps part of the reason his Greek education failed was the emphasis on things he found uninteresting. The extended grammar drill necessary for Greek detracted from the pleasure of the great literary texts. He also complains of the way his teachers punished and threatened him, which kept his affective filter high. He acquired Latin naturally as a first language and though not fond of those who taught him his letters (i.e. to read and write Latin), he was fond of those grammarians who worked to teach him the Latin literary language: adamaveram enim latinas, non quas primi magistri, sed quas docent qui grammatici vocantur (‘for I had fallen in love with

Latin, not that which the first schoolmasters teach, but that which those teach, who are called grammarians teach’)(1.13). He most enjoyed those who worked to teach him the best Latin through imitation of the great authors, which was the method of the rhetorical school, et maior iam eram in schola rhetoris et gaudebam superbe (‘and now I was older in the school of the rhetorician and I rejoiced proudly’) (3.3).

Teachers and grammarians added commentaries to the classical Latin texts used in schools.

Donatus’ own commentary on the plays of Terence worked to make texts from an earlier period of Latin understandable. He added glosses of the unusual forms and references. A large portion of his commentary, however, is related to gesture and the staging of the performance, through which he attempts to supply for his students the visual element of performance, which would have aided comprehension (Demetriou 2014: 8).

Oakes 30

Whether or not Donatus, Servius,40 or the other early grammarians were writing textbooks to teach Latin as a second language, their works were certainly used as such in later periods.41 During their lifetimes,42 however, speakers of vernacular Latin, what would become the Romance languages, would have struggled to understand the poetic language of Virgil or Terence, or the abundantly rhetorical style of Cicero; nonetheless, they were still capable of understanding basic Latin or a simple sermon

(Leonhardt 2013: 118).

In 999 CE, Pope Gregory V died and his epitaph represents the “earliest extant text which can be interpreted to imply a distinction between the language spoken by the populace of Rome and Latin”

(Clackson and Harrocks 2007: 267). From this point on, the pedagogy of Latin in Western Europe changed greatly. Where previously grammarians aimed to teach a “fixed” literary form of their students’ first language, they may now be understood to be teaching a second language.

The High Middle Ages and Communicative Latin Language Learning (10th-12th Centuries CE)

The learning of Latin in northern Europe was part of a Christianized idea of education, from

Augustine, Cassiodorus,43 Isidore,44 and Bede.45 Bede based his grammar, as many of his predecessors

40 Servius’s commentary on the Aeneid (ca. 400-420 CE) “reflects the teaching practice of his time: it proceeds on a word-by-word and line-by-line basis. While the commentator offers a panoply of information, including facts about philosophy, history religion, and the sources of literary allusions, his primary focus is language. Here he devotes special attention to more technical linguistic and rhetorical aspect of language use rather than literary or aesthetic matters. His other main focus is the function of the poet as a potential model for student” (Copeland and Sluiter 2009: 126). 41 See the following three sections, pp. 31-50. 42 Leonhardt 2013: 118 sets the earliest date for when the vernacular separated from Classical Latin at approximately 600CE. 43 Cassidorus wrote the Expositio pslamorum (ca. 540) and Institutiones (ca. 562) as well as other spiritual writings: Copeland and Sluiter 2009: 210. 44 Isidore of Seville (ca. 625) wrote the Etymologiae, an encyclopedic work. In the first book, De grammatica, he “combines a technical approach to grammar with a more mystical view on the value of letters as an indication of moral truth” (Copeland and Sluiter 2009: 234). 45 Bede wrote De arte metrica and De schematibus et tropis (ca. 720) for the purpose of explaining meter and grammar from a Christian perspective: Copeland and Sluiter 2009: 256.

Oakes 31 and contemporaries did, on the earlier grammars of Donatus, Priscian,46 et alii.47 This tradition of grammars, as far as teaching the language was concerned, focused on morphology and included only the syntax necessary to put words into sentences or understand Latin orthography.48

Alcuin of York (ca. 731-804 CE), called Albinus in the court of Charlemagne, promoted reforms which brought a new style of teaching Latin as a second language into prominence. This method built on those of earlier periods, but systematized instruction to distinguish Classical Latin from the Romance dialects that were starting to be recognized as distinct languages. Charlemagne and his father Pepin realized that both the Latin used by priests for their observances and the Latin in the texts they were referencing were filled with errors (Leonhardt 2013: 122-62). This led to the language reforms of

Charlemagne, for which Alcuin was a guiding figure.

Alcuin wrote his own Ars grammatica49 using content mostly from Donatus and Priscian, but made a substantial change from their methods of framing the grammar content. He continued the question and answer format of Donatus, but breathed new life into it. Alcuin’s grammar follows the same series of topics (voice, letter, syllable, parts of speech …) and question and answer format as Donatus’ Ars major, but enlivens it substantially through the introduction of student characters who add personality and humor to the conversations:

Fuerunt in schola Albini magistri duo pueri, unus Franco, alter Saxo, qui nuperrime spineta grammaticae densitatis irruperunt. Quapropter placuit illis paucas litteralis scientiae regulas memoriae causa per interrogationes et responsiones excerpere. At prior illorum Franco dixit Saxoni: Eia, Saxo, me interrogantem responde, quia tu majoris es aetatis. Ego XIV annorum; tu ut reor XV. Ad haec Saxo respondit: Faciam; ita tamen, ut si quid altius si interrogandum, vel ex philosophica disciplina proferendum, liceat magistrum interrogare. Ad haec magister: Placet, filii, propositio vestra: et libens annuo vestrae sagacitati. Et primum dicite unde vestram convenientius disputationem esse arbitramini incipiendam?

46 Priscian (ca. 250) wrote several grammars aimed at native speakers of Greek who were already competent in Latin: Copeland and Sluiter 2009: 168. His work is discussed on p. 38. 47 See Copeland and Sluiter 2009: 62-299 for further information on the early grammarians. 48 Other moral and philosophical interests were included. See n. 45 on Isidore. 49 The Latin text cited in Copeland and Sluiter 2009: 272- 98 is from Migne PL 101.

Oakes 32

Dis. Unde, domine magister, nisi a littera? – Migne PL 101:853B

There were in the school of the teacher Albinus two boys, one a Frank, the other a Saxon, who very recently broke into the thorniness of grammatical density. Therefore, it pleased them to pick out a few rules of the letter-related knowledge for the sake of memorizing them through questions and answers. But the elder of those, the Frank, said to the Saxon: “Hey, Saxon, you answer me when I am asking because you are older. I am fourteen, you, as I think, fifteen.” To these, the Saxon answered: “Sure, as long as if something deeper must be asked or explained from the philosophical discipline, it is permitted to ask the teacher.” To these things, the teacher said: “Your idea is pleasing, sons: and happily, I agree to your keenness. And tell me first from where you judge your disputation should be more agreeably begun?” Students: “From where, master-teacher, unless from the letter?”

Alcuin explicitly states that his students are embarking on a thorny subject, which may indicate a desire either to empathize with his students or to challenge them. The method employed is memorization through question and answer. Memorization is a feature of learning, but the kind of classroom Alcuin is modeling is one open to acquisition. The conversation is relaxed and the teacher is there to shelter the students by stepping in when necessary. A less advanced student is getting help from a more advanced student, aiding i+1.

As the conversation proceeds, the younger boy asks questions, the older boy provides answers, and the teacher explains, and muses further upon, the philosophy of what they are discussing.50 There are asides and diversions. When the teacher’s discussion of the syllable begins to stray too far into meter, he cuts himself off, and suggests that they begin with the parts of speech (Migne PL 101: 857D). The students also stray into a side discussion on what grammar actually is, and this discussion forms a sort of table of contents (Migne PL 101: 857D-858D). The teacher tells his students, Vestra curiositas modum non habet. Ideoque modum manualis libelli excedere vultis (‘Your curiosity has no limit. Therefore, you wish to exceed the limit of this little handbook.’) (Migne PL 101:858D). The diversions into philosophy

50 Vocales sunt sicut animae, consonantes sicut corpora. Anima vero et se movet et corpus. Corpus vero immobile est sine anima. (“Vowels are like souls, consonants like bodies. For the soul moves both itself and the body, but the body is immobile without the soul.”) Migne PL 101: 855B

Oakes 33 represent the understanding of the importance of moral learning and the need for Latin to access it.

There are also times where the boys quote ancient authors, such as Virgil. Alcuin is mixing the future content the boys are meant to learn in Latin with their education in the language, a form of either pre- teaching,51 or maintaining motivation by reminding students of the more meaningful material that lies ahead.

Unlike Donatus’ simple grammar and lists of properties for memorization, Alcuin’s style mimics an ideal schoolroom with curious students who use whatever grammar is necessary for the conversation.

Beyond simply providing examples, he provides explanations for the terminology:

Fa. Nunc, ut estimo, ordo accidentium nominis poscit a generibus inquirere. Unde genera dicta sint, primo velim scire. – Saxo. Genera dicta sunt a generando. Quapropter proprie genera sunt, quae generare possunt: id est, masculinum, femininum. Nam neutrum et commune magis vocis qualitate, quam natura dignoscuntur. Epicoena promisca sunt, et sub una voce profertur masculinum et femininum. Fa. Quid est inter commune et epicoenum? – Saxo. Quod commune modo masculino, modo feminino articulo, prout constructio orationis eget, iungitur. Epicoenon semper uno articulo profertu, et utriusque naturae animalia una voce et articulo significat. Fa. Num omnia nomina certi sunt generis? —Saxo. Nequaquam. Fa. Unde sunt dubii generis nomina? Forte differentiae causa? – Saxo. Non differentiae, nec aliqua necessitatis causa, sed sola auctoritate veterum, ut: hic et haec finis, cortex, silex, … Migne PL 101:862A-B

Fr. Now, as I judge it, the order of the properties of the noun demands enquiring about the genders. I’d like to know first from where the genders are named. – Saxo. The genders are named from generation. Therefore, particularly, genders are able to generate: that is, masculine and feminine. For the neuter and common are more known by quality of voice than by nature. The epicene are indiscriminate, and are expressed under one voice, masculine and feminine. Fr. What is [the difference] between common and epicene? – Saxo. That a common noun is joined sometimes with a masculine, sometimes with a feminine article, as the construction of speech requires it. An epicene noun is always found in one voice, and in one voice and article signifies creatures of each nature. Fr. Surely not all nouns are of certain gender? – Saxo. By no means. Fr. From where are nouns of doubtful gender? Perhaps on account of difference? – Saxo. Not of a difference, and not in some other cause of necessity, but by the authority of the ancients alone, as: hic et haec finis, cortex, silex, [‘this masculine and this feminine end, husk, stone’] …

51 Pre-teaching is another form of sheltering instruction. It builds a base for second language students to rely on during the content lesson. See Buehl 2014: 12-17.

Oakes 34

The grammar Alcuin’s students use in the text is more complex than that of Donatus, but also includes repetitions of the same information in simpler terms, which makes it more comprehensible than

Donatus as well. Speculative, logic-based grammar is represented (e.g., explaining that “genders are able to generate”), but importance is placed on learning grammar in context (e.g., explaining that some nouns are of doubtful gender and must be determined solely by how the ancients used them).

Regarding common and epicene nouns, the Frank asks the obvious questions, searching for rules to help discern the categories, only to be told that he must rely on the authority of the ancients and memorize/internalize the genders.

Alcuin lowers his students’ affective filter by introducing humor into the text. He pre-teaches future content by mixing in philosophical or religious material that is relevant to the students. He uses repetitions and examples to make it comprehensible. These adaptations are a form of sheltering instruction, and they suggest a pedagogue rather than a grammarian. While the two professions are often mixed, Alcuin provided no innovations of grammar in his text, only a new method of presenting it.52 His method is more communicative than the earlier methods. He attempts to lower the affective filter, provide meaningful content, and provide repetition/alternate explanations to facilitate comprehension. His students were expected to be fluent in Latin by acquisition of the language.

Alcuin learned and taught the language as part of a Christian system of education. This system in some ways differed very little from the Roman education plan described by Quintilian (both for the first language and for learning Greek) (Murphy 1980: 162). Both systems started with learning how the sounds relate to the letters, then, with the aid of a simple grammar, advanced to learning how letters form syllables and words. Soon phrases and constructions were introduced, usually through dictation

52 The other great innovations credited to Alcuin are his strong support of a standardized orthography, Carolinian miniscule script, and punctuation. His own texts were thus highly readable in comparison to many before them. He also helped reform the system of copyists in monasteries, by requiring that scribes should read what they were copying and focus on getting the text right: Duckett 1951: 122, 256-61.

Oakes 35 and memorization of a book of simple phrases deemed useful for students, e.g. what it means to be a good student or collections of the dictums of great authors. Then the students learned connected grammar through stories like the fables of Aesop. At this point the students slowly began to read the auctores or great authors,53 and learned to model them.

In Christian monasteries such as Alcuin’s at York, students were surrounded by spoken Latin through the daily rituals and ceremonies of the Church. They heard and distinguished the sounds of

Latin before they studied a prescriptive grammar. Motivation to learn was high because of the need for

Latin in their observances. The simplest phrase reader used since late antiquity was the Disticha Catonis, ethical injunctions in hexameter couplets (Thompson and Perraud 1990: 49-85). Christian schools often substituted the Psalter (Black 2001: 37). Rather than Aesop’s fables, many religious texts were used by intermediate readers. These texts included those of the Latin church fathers, Bede’s history, and the many lives of the saints which were being published. Christian texts of moral instructions superseded classical non-Christian texts. This indicates a pedagogical focus on content over grammar/eloquence.

Composition was practiced through imitations and transformation exercises: summarizing memorized colloquia, writing a story in a specific style, making prose from poetry or vice versa, translating a Greek rhetorician into Latin or vice versa (Murphy 1980: 162). Following the ars grammaticae,54 many students training to be secretaries or notaries learned the ars dictaminis or principles of composition (Grendler 1989: 115). These positions enabled correspondence between the church, governmental, or provincial authorities. The practice of writing letters was very formalized and numerous examples of different kinds of letters, written to be imitated rather than culled from classical

53 Different ideas of who were acceptable authors for imitation existed in the different schools and periods. The Aeneid was in all of them. See Black 2001, the entire book, for which texts were used when and where. He includes charts explaining how many manuscripts of each text are evidenced when and where towards the end of each chapter. 54 This was not only the name of numerous grammar textbooks, such as Alcuin’s, but also the name for the lowest stage of learning.

Oakes 36 authors, were provided for imitation. Principles of prose composition were taught according to Cicero’s early work on oratory, De inventione, and the pseudo-Ciceronian Rhetorica ad Herrenium (Grendler

1989: 115).

The medieval program of education, starting with distinguishing sounds and moving to original compositions, aligns well with Krashen’s plans for second language learners. The educational program began with simple grammar teaching, advanced through sheltered activities, and culminated in content area work. The language is practiced in multiple ways: dictation, memorization, recitation, interrogation, transformation (listening, speaking, reading, writing). In most cases, students are listening to and interacting with fellow students who are learning at levels lower than, the same as, or higher than their own, which can only be helpful for i+1 acquisition. The classroom was only as bilingual as it needed to be for understanding. The series of composition activities was graded in difficulty, but the grading related to lower vs. higher order thinking and relatable content vs. more abstract content. It is not graded by the difficulty of grammar. Grading for grammar ignores the natural order and i+1 hypothesis, which results in either forcing structures before the student is ready or failing to provide the structures they are ready to acquire. Grading for content respects a student’s cognitive ability to interact with that content and helps build background that can shelter more difficult content when it is introduced. The content is also meaningful (mostly moral and religious) and integrated into language learning through the choice of texts.

This medieval program, ars grammaticae, auctores, and ars dictaminis, was a time-consuming process with a great deal of dictation and memorization (partly imposed by the lack of copies of the texts), but produced some exemplary Latinists, some educated enough for the literary transactions of life, and others with the basic ability to read.55 Acquisition was always the goal and motivation was

55At the time (late antiquity until the rise of the vernacular in the fourteenth century) in Europe, Latin literacy was literacy. It is not that written texts in the vernacular languages did not exist, but rather that Latin was the lingua franca. Aelfric’s English grammar (999 CE) is known as the first grammar of Latin not written in Latin (Menzer

Oakes 37 sustained by the need for Latin in the church, legal system, and most literary endeavors. Grammars were reserved for early learners. Intermediate learners read the auctores. Finally, at the higher levels, the use of Latin as the only tool for communication maintained and expanded Latin skills.

The Later Middle Ages and Rise of the Vernacular (13th-14th Centuries CE)

This period in some ways represents a step back for Comprehensible Input style instruction since the focus moved more towards technical grammar and use of the vernacular to aid comprehension. Alcuin’s Ars grammatica represented an effort to combine Donatus’ Ars maior with

Priscian’s Institutiones grammaticae. Priscian was a native-speaker and teacher of Latin in the sixth century who taught adult learners in Greek-speaking Constantinople. His grammar differed from that of

Donatus in multiple ways; the most important distinction was that Priscian’s grammar included two books on syntax (XVII and XVIII), which made up at least a fourth of the work (Copeland and Sluiter

2009: 169). Donatus’ Ars maior includes syntax only incidentally (Copeland and Sluiter 2009: 83). These syntax portions and creative use of grammatical metalanguage (Copeland and Sluiter 2009: 168) make it one of the first “speculative grammars,” which attempted not only to describe what acceptable grammar was, i.e. what could be discovered through exemplars, but also to derive the rules of grammar.

Priscian was “trying to improve on traditional Latin grammar by applying Greek theoretical

(philosophically inspired) notions” (Copeland and Sluiter 2009: 168). Because Priscian taught adult learners already familiar with the morphology of Latin and literate in their first language, his grammar is technical and could not have been used as a beginner’s text. Priscian also produced a text called the

Partitiones based on an older Roman method of teaching in which he parsed the first verse of each book of the Aeneid (Black 2001: 45-6).

2004). Latin was the only subject taught in primary schools because it allowed access to everything else a student might wish to learn later. Its nature as a fixed language aided it tremendously as the only language at the time with standardized spelling and grammar. Grammar was assumed to be the same in all languages (Thomas 1995: 345) and learning Latin grammar was also learning what was needed for the first language.

Oakes 38

Despite the advanced nature of Priscian’s texts, adaptations for early learners of Priscian’s speculative and parsing grammars multiplied in Italy (Black 2013: 62-63). Ianua56 was a parsing grammar. It followed Donatus’ model of the parts of speech and a question and answer format, but the questions were based on parsing a specific word and employed the speculative, usage-based terminology generated by Priscian rather than the exemplar method of Donatus: Poeta que pars est?

Nomen est. Quare? Quia significat substantiam et qualitatem. (‘Poet, which part of speech is it? It is a noun. Why? Because it signifies the nature and quality’). The questions cover a complete parsing and morphology (Grendler 1989: 180-81).

The Ianua came to be used as a basic first reading text. It was read a first time for pronunciation and again for pronunciation and meaning (perhaps with teacher-provided interlinear translation)

(Grendler 1989: 184). The next reading concentrated on memorization and discussion. While the repetition aided in comprehension, the content was entirely grammar-focused. In order to attain i+1

(the level of complexity that is next on the natural order rather than the next level of grammar which the teacher expects the student to learn), the teacher must routinely check for comprehension and provide alternate input until comprehension is achieved. Parsing grammars like the Ianua use terminology disconnected from real life and the repetition with re-reading of a single grammar does little to make that terminology comprehensible. Ianua is poorly conceived to serve as an introductory text when measured against Krashen’s hypotheses.

After basic instruction with a text like the Ianua, other grammars based on the parsing method were written to serve as secondary textbooks (replacing the simple readers). Alexander of Villedieu’s

Doctrinale (1199 CE), a reinvention of Priscian’s speculative grammar and the Ianua, was written in verse to make memorization easier, but required commentaries and vernacular glossing to make it

56 So named for the first word of its verse prologue: Ianua sum rudibus primam cupientibus artem (‘I am the door for the uneducated who desire primary instruction’): Grendler 1989: 176

Oakes 39 comprehensible (Black 2001: 52, 74-80). The commentary tradition filled the gaps between memorization (learning) and understanding (acquisition). Secondary prose textbooks or summe57 were also developed (Black 2001: 94). While in some schools early reading in the Ianua was still conducted entirely in Latin, the secondary school summe were designed for working with the vernacular and included translation exercises called themata and Latin/vernacular vocabulary lists (Grendler 1989: 196-

97). The summe represent an early form of Grammar-Translation that was moving away from

Comprehensible Input style instruction.

A division between acquisition and learning also existed in the teaching of composition. There came to be a strict division, in effect two dialects of Latin: ordo naturalis and ordo artificialis. Ordo artificialis was only taught after mastery of ordo naturalis. As pupils were first taught to write, they followed first language word order; the more eloquent and rhetorical classical word order was for high register writing (Black 2001: 337). The ordo naturalis was the result of a desire for a middle ground between the basic grammar school letters and reading, or the longer, more detailed study of the auctores. The ordo naturalis was a learner language, full of medieval “barbarisms,” but one that was effective for communication, as clerks, scribes, and businessmen used Latin for creating various types of contracts and writing letters. That the ordo naturalis was considered as an acceptable form of the language, though a lower register, demonstrates an acceptance of learner languages. When comprehension was the priority, there was no need to introduce artificial corrections that activated monitor learning and impeded acquisition.

Aside from grammar and writing instruction, by the end of the thirteenth century there was a fairly well developed collection of reading texts, with glosses and vernacular vocabularies attached,

57 Summe/summae (pl), Summa (sg.). Summa was often the first word used in the title of these secondary school prose textbooks. For example: Pietro de Isolella’s Summa grammaticae (Summation of Grammar): Black 2013: 88.

Oakes 40 called the auctores octo: Cato’s Disticha,58 Theodulus’ Ecloga, Facetus, Chartula, Liber parabularum,

Tobias, Aesopus, and Floretus59 (Grendler 1989: 111-14). There was no strict ordering to the texts, but they were joined, two or three to a manuscript, and circulated together. Many more texts were substituted or included in earlier versions. They included Avianus’ Fabulae, the elegies of Maximianus,

De raptu Proserpinae by Claudian, Statius’ Achilleis, Boethius’ Consolatio philosophiae, and other texts by late antique and early medieval authors. Boethius’ text served as a transition for intermediate readers, with copious grammar structures and references to the classical authors who would follow, primarily Virgil, Ovid, Statius, and Lucan (Grendler 1989: 114). This grading of texts represented a desire to start with something familiar in styles or context, anticipate the content to come, and finally reach the unfamiliar. As Krashen’s hypotheses require comprehension at each level before moving on, it is important that not just the grammar, but also the content be comprehensible. Notice, however, that, as each stage of the reading curriculum intermixes poetry and prose; poetry was not considered a separate topic or stage of Latin teaching.

This period of teaching was transitional. In many places, particularly monasteries, Latin teaching followed the same patterns established in the Carolingian Renaissance, but with a growing understanding of the artificiality of classical texts and increased use of medieval ones. The acceptance of learner languages and more contemporary texts marked a positive shift towards a communicative pedagogy. In some places, however, new early reading texts were adopted that employed incomprehensible grammar. Rather than return to older patterns, teachers kept the speculative

58 These were not actually written by Cato, but compiled in late antiquity and attributed to him because of a well- known tradition that he had written such a book for his son: Thomson and Perraud 1990: 51. 59 In the Ecloga, characters engage in a poetic contest with stories from mythology and the Old Testament. The Facetus is a verse manual of good manners. Chartula is the first word of De contemptu mundi, a verse treatise, but became the more common name for the text. The Liber parabulum is a book of verse proverbs; Tobias is a verse morality poem; Aesopus is the collection of Aesop’s fables; and Floretus is a religious poem.

Oakes 41 grammar and added heavy memorization, the vernacular, and translation-based methods to make the speculative grammars comprehensible. In the Renaissance, there was a strict return to classical texts.

The Renaissance and the Humanists vs. the Practicalities (15th-16th Centuries CE)

The humanist movement developed from the Renaissance glorification of classical ideals. It was not simply a return to those forms, but a belief that they could improve one’s character. Studying the classical ideals would make one a better person and a better leader. In 1402/3, Pier Paolo Vergerio advocated for teaching according to humanist methods in this way (Grendler 1989: 117-18):

Nam liberalibus quidem ingeniis, et iis qui in publicis rebus hominum communitate versari debent, convenientiora sunt historiae notitia et moralis philosophiae studium. Ceterae quidem enim artium liberales dicuntur, quia liberos homines deceant, philosophia vero idcirco est liberalis, quod ejus studium liberos homines efficit. In horum igitur altero praecepta, quid sequi quidve fugere conveniat, in altero exempla invenimus: in illa enim omnium hominum officia reperiuntur, et quid quemque deceat, in hac vero quid factum dictumve sit suis quoque temporibus. Adiciendum est ad haec (ni fallor) et tertium, id est eloquentia, quae civilis scientiae pars quaedam est. Per philosophiam siquidem possumus recte sentire, quod est in omni re primum, per eloquentiam graviter ornateque dicere, qua una re maxime conciliantur multitudinis animi; per historiam vero in utrumque juvamur.60

For indeed to those with liberal minds, and those who ought to be engaged in public affairs and the community of men, more useful are an acquaintance with history and the study of moral philosophy. For indeed others are called liberal arts, because they are fitting for free men, and in truth philosophy is liberal, because the study of it creates free men. Therefore, in one of these, we find lessons, as to what is fitting to follow or what to avoid, in the other we find examples: for in the former the duties of all men are obtained, and what befits each; in the latter, however, what was done and what was said in their own times too. It must be added to these (unless I am deceived) also a third, that is eloquence, which is also a certain part of the civil knowledge. Through philosophy accordingly we are able to understand correctly which is the first thing of all; through eloquence to speak seriously and ornately, by which one thing the souls of the multitude are most won over; through history, on the other hand, in both pursuits we are aided.

This argument for the studia humanitatis, that they prepared men both to live their lives well and to engage in civil life, was a strong motivation for the Renaissance. Humanists found no value in studying post classical Latin.

60 Latin text from Robey 1980: 30.

Oakes 42

Speculative grammar was set aside, its brief heyday ended as grammar studies were deemed suitable while learning to read, but not as an important study in their own right. A humanist grammar by

Niccolo Perotti was created as a comprehensive school text (Grendler 1989: 173-74). His Rudimenta gramattices (1473 CE) included the primer material of the alphabet and prayers, and a question/answer section of pedagogical principles and grammar followed by a treatise on letter writing. He mixed his grammar explanations with sections borrowed from earlier grammars and provided examples of both artificial sentences and those from ancient authors. The grammar attempted, however, to excise all constructions and orthography introduced by medieval writers. The development of new humanist grammars continued to flourish. Parents and city councils, however, still expected their children to learn as they had learned, from Donatus and the Disticha,61 and the new texts were not used extensively

(Grendler 1989: 118). The humanist’s greatest curricular result was the replacement of all the medieval early-readers with Cicero, Virgil, and other classical authors.

In the Renaissance, the speculative grammar, etymologies, and early moralistic reading texts were put aside in favor of classical sources. Particularly the ars dictaminis, the composition books based on theory and imitable examples, seemed too removed from the best classical style. While Cicero had been seen as the model in letter writing, few of his works were actually in use.62 Between 1392 and

1421, however, many lost manuscripts of his works were found along with more complete texts of those few already in use. These included his rhetorical treatises De Oratore and Brutus. In some schools,

Cicero’s letters served as early readers, his speeches as intermediate readers, and his rhetorical texts were used for composition instruction (Grendler 1989: 405). A whole curriculum of Cicero could be followed after students had learned basic reading skills through letter primers,63 such as the Ianua, and

61 Colloquia also began to be developed to support early learning. These are discussed on p. 47. 62 See p. 37. 63 I.e., Carta (letter cards) and tavola/tabula (‘hornbooks’ or wooden tablets with the letters visible under a thin, see-through plate of horn) had examples of letters, syllables (ba bo bi bu), and simple prayers, primarily the Pater

Oakes 43 the Disticha Catonis, or the Psalter. Ciceronian eloquence was now the sine qua non of a broad range of civic and ecclesiastical pursuits.

Numerous editions of classical texts were produced besides those of Cicero, but many lacked the long medieval tradition of grammatical, rhetorical, etymological, or allegorical commentary.64

Instead, these new texts were produced as school texts, with paraphrase commentaries containing the information required to read and understand the text literally. These paraphrase commentaries were an important pedagogical tool and anticipated Comprehensible Input style instruction. The practice had presumably been used before in medieval pedagogy by teachers seeking to make texts comprehensible for students, but now was formalized. Whereas previously the formalized knowledge included with texts was from the commentary tradition, now the paraphrase commentaries were provided in the text.

These difficult classical texts had to be adapted for early reading and that is the place of the paraphrase commentary (Grendler 1989: 244-49). For example, Grendler (1989:245) supplies a Latin paraphrase for the first three lines of the Aeneid: Arma virumque cano Troiae qui primus ab oris/ Italiam fato profugus, Lavinia venit/ litora … (‘I sing of arms and the man who first exiled from the shores of

Troy by fate came to Italy and Lavinian shores ...’):

Ego Virgilius, cano, id est canto, arma, id est bella, que, pro &, virum, scilicet Aeneam illum hominem fortissimum, qui, id est Aeneas, primus, id est ante alios, profugus, it est exul, fato, id est deorum praeordinatione, venit, accessit Italiam, pro in Italiam, ab oris, id est a regionibus, Troiae illius provinciae, & civitatis in Phrigia.65

I Virgil, I sing, that is I recite, arms, that is wars, que in place of “and,” a man, clearly that most brave man Aeneas, who, that is Aeneas, first, that is before others, a fugitive, that is an exile, by fate, that is by the preordination of the Gods, came, reached Italy, for into Italy, from the shores, that is from the regions, of the province Troy and a city in Phrygia.

Noster, which were meant to be pronounced and traced as students first learned to associate sounds and letters: Grendler 1989: 142-61. 64 Medieval scholars had little interest in history, for example, and had almost entirely ignored Livy: Grendler 1989: 255. 65 Grendler found this paraphrase quoted in a dialogue which “discussed pedagogical practice in primary and secondary schools.” His source is Bonamico and Paleario (1567): Il grammatico.

Oakes 44

While this formalized, repetitive language is uninteresting to one who already understands the text, the repetition, Latin synonyms, and paraphrasing made it comprehensible for students. These commentaries also included background content and vernacular grammar explanations. Using alternate

Latin glosses for every word in a paraphrase is a simple way to help comprehension. In previous commentaries meant for higher levels, glosses had only been provided for difficult words and were intermixed with copious other types of scholia.66 Provided the teacher was checking the students’ comprehension (perhaps by asking them to summarize) before moving onto the next portion, this type of lesson could easily succeed in a modern Comprehensible Input classroom. The commentary helps teach vocabulary in context and focuses on understanding meaning.

The great classical poets, such as Virgil, Horace, Terence,67 and Ovid, who had been less studied in the medieval classroom, were given similar commentary treatment by Giovanni Fabrini (1516-1580).

He also produced an edition of Cicero’s Epistulae ad familiares, which was used in the classroom to support the Renaissance dictum that one should write like Cicero. (Grendler 1989: 247)

In the Renaissance, learning to write like Cicero was a more complicated composition process than that practiced in medieval classrooms. Schools established by the Jesuits in cities across Italy and

Southern Europe,68 modeled after the one founded by the order in Messina in 1547, greatly stressed prose composition motivated by competition. Competition was woven into the entire college and there was little concern for the students who always lost (Grendler 1989: 381). They were bastions of active

Latin use. They eschewed the vernacular and were imaginative in their exercises, e.g. staging Latin plays or poetry events. An interesting feature of the Jesuit schools was that, while they had a strict humanist curriculum and graded classes, students were not promoted by age or completion of the class, but when

66 Cf. Servius’ commentaries, n. 41. 67 Terence was the text of choice for teaching conversational Latin, although conversational register Latin was increasingly being supplanted by Ciceronian high style. See p. 48. 68 While Jesuit schools might differ in their patrons or students, “all of the Jesuits taught the same texts and authors in a methodical progression of classes.” Grendler 1989: 379.

Oakes 45 they demonstrated mastery. The Jesuits’ goal was for students to be able to use Latin as if it was their first language. (Grendler 1989: 377-81)

The first step in learning to write like Cicero was to read Cicero, whose style is dense and marked by extensive subordination. Rather than teach students to listen to or read his speeches in Latin word order,69 students rearranged the words into an order that paralleled first language meaning. They were taught to assign letters to each word so that they could move from word A to word B, and so on

(Grendler 1989: 223). After reading some of Cicero’s letters, teachers and students analyzed them and tried to assign them to categories such as apologetic, celebratory, and persuasive. Attention was paid to the register of words used (especially which were high style, which were low), what words came in phrases (collocation), the emotional force of specific words, and which words contributed to Cicero’s prose meter (Grendler 1989: 225). The teacher then assigned themes (writing exercises with a specific topic, audience, and purpose) and the students imitated Ciceronian style as they wrote their own versions. These themes were corrected by the teacher and returned (Grendler 1989: 232). Students do not seem to comprehend the language as language, which would lead to acquisition, but instead are involved in decoding and imitating, which results in learning. According to Krashen’s hypotheses, the

L1+monitor is engaged when a task is beyond a student’s acquired capability in the language and a concern for correctness is what initiates the monitor. Thus, while this type of prose composition activity may reinforce some already acquired language skills, the new content provided is being learned consciously through the monitor. The humanist overvaluing of Ciceronian style was famously criticized by Erasmus.

69 Consider that old classicist’s joke: This version is from the LatinStudentProblems twitter: “A senator was 15 minutes late to Cicero's speech at the Senate. He sat down & quietly asked another senator what Cicero was talking about. The other senator replied, ‘I don't know. He hasn't gotten to the verb yet!’” https://twitter.com/tweetsfromlsp/status/677587020076572673

Oakes 46

Erasmus was one of the Renaissance writers of colloquia. We have seen above that this use of dialogue as a teaching tool was popular in antiquity; as an instructional model, it is reinvigorated in the

Renaissance. Juan Luis Vives (1492-1540) wrote the Colloquia sive linguae Latinae exercitatio (‘Colloquia or an Exercise of the Latin language’), which was sometimes used in Italian grammar schools either as replacement for or to supplement the Disticha Catonis (Grendler 1989: 200-01). These dialogues closely resemble those found in the schoolbook portions of the Hermeneumata. There are twenty-five of varying lengths and they present scenes from the life of an upper-class schoolboy. They contain topics familiar to schoolboys, and age-appropriate humor.

Heri vidi scelus capitale admitti. Ludimagister vici recti capro olidior, qui in phrontisterio suo in paedore et foetore docet auditores diobolarios, ter aut quater pronuntiavit volucres accentu in paenultima; ego vero miratus sum terram illi non dehiscere.70

Yesterday I saw a capital crime committed. The schoolmaster of Straight Street, smellier than a goat, who in his own school, in the filth and stink teaches his two-silver-obol students, three or four times pronounced volucres with the accent on the penult; I truly marveled that the earth did not split open for him.

There is much good comprehensible input Latin found in these colloquia. Their content is interesting to a young boy and the humor helps lower the affective filter. There are also references to classical texts which prepare the boys for the next stage of learning.

Erasmus was a priest, scholar, and teacher who studied in Paris but traveled extensively and interacted in the international community of humanists. He wrote two, important early teaching manuals On the Method of Study and of Reading and Interpreting the Authors (De ratione studii ac legendi interpretandique auctores, 1511 CE) and On Educating Children at once and Instructing Liberally

(De pueris statim ac liberaliter instituendis, 1529 CE), and many colloquia along with numerous theological works (Rummel 1990: 5).

70 Johannes Ludovicus Vives, Exercitatio Linguae Latinae, 7- Refectio Scholastica. (1782), From Perseus Tufts Colloquia Collection accessed via Stoa.org: http://www.stoa.org/hopper/text.jsp?doc=Stoa:text:2003.02.0010

Oakes 47

In On the Method of Study,71 Erasmus lays out a typical humanist education plan, but emphasizes the importance of treating Latin as a language to be acquired and used creatively. He considers Terence as the author that should be most imitated:

Rursum inter latinos quis vtilior loquendi auctor quam Terentius? Purus, tersus et quotidiano sermoni proximus, tum ipso quoque argumenti genere iucundus adoclescentiae. (ASD 1.2: 115-16)

Again, among the Latin writers who is more useful an authority of speaking than Terence? He is pure, brief, and nearest to everyday speech; then, also by the very nature of his subject, he is pleasing to youth.

Erasmus’ teaching philosophy advocates respect for the interests of his students and for keeping them positively engaged in the lessons. In On the Education of Children,72 he promotes a teaching philosophy that discourages corporal punishment and argues for the importance of play. He stresses the importance of lowering the students’ affective filter, though he would not have used that term. Yet, he demonstrates an understanding of how the affective filter works in the classroom, when he writes about teachers who punish as opposed to teachers who encourage and reward:

Nihil enim est inutilius quam quum praeceptoris mores efficiunt vt prius odisse incipiant studia quam intelligere possint quare sint amanda. Primus discendi gradus est praeceptoris amor. Progressu temporis fiet, vt puer qui prius literas amare coeperat propter doctorem, post doctorem amet propter literas. Nam quemadmodum munera pleraque vel hoc nomine gratissima sunt, quod ab his proficiscantur quos egregie charos habemus, ita literae, quibus nondum iudicio placere possunt, his tamen doctoris affectu commendantur. Rectissime dictum est ab Isocrate, eum plurimum discere, qui discendi cupidus est. (ASD 1.2: 53-54.)

For nothing is more useless than when the character of the teacher brings about that his students begin to hate their studies sooner than they can understand why they are to be loved. The first step of learning is love of the teacher. With the advancing of time it will happen that the boy who first began to love his letters on account of his teacher after loves the teacher on account of the letters. For just as many gifts are especially or for this reason most pleasing, because they are bestowed by those whom we hold especially dear, thus the letters, to those whom they are not yet able to be pleasing on account of their own judgement, to these nevertheless they are commended by their affection for their teacher. Most right is that dictum from Isocrates, “He learns the most who is desirous of learning.”

71 Thompson 1978: 661-691. Latin text from ASD (Amsterdam) 1971: 111-51. 72 Rummel 1990: 65-121. Latin text from ASD 1971: 24-78.

Oakes 48

Erasmus subscribed to the belief that language learning is easiest for children, but recognized that the learning program had to be suitable for children. He argued for the use of appealing texts, like the fables of ancient authors and the use of pictures to make them both more interesting and more comprehensible (ASD 1.2: 66-67), Nihil autem vetat quo minus voluptati comes sit vtilitas et iucunditati iuncta sit honestas. Atque haec omnia tam frugifera nullo taedio discit puer. (‘Nothing however forbids usefulness as a companion for pleasure and integrity [as a companion] for agreeableness. And the child learns all these very fruitful things with no tedium.’) (ASD 1.2: 69)

Erasmus criticized harshly those humanists who praised Cicero above all others. In his Ciceronian

Dialogue (Dialogus Ciceronianus),73 he criticizes a Ciceronian for speaking stylishly, but without content:

Tam Romanae dixit Romanus iIle, vt nihil audirem de Morte Christi. Et tamen ilIe Ciceronianae dictionis ambitiosissimus candidatus, Ciceronianis videbatur mirifice dixisse, quum de re pene nihil diceret, quam nec intelligere, nec amare videbatur, neque quicquam apposite dicebat, nec vIlos mouerat affectus. Tantum hoc laudis ferebat, quod Romane pronuntiasset, et aliquid Ciceronis retulisset. Probari poterat hoc velut indolis ingeniisque specimen, si a puero apud pueros in schola fuisset habita talis oratio. Verum ad talem diem, ad tales auditores, ad tale argumentum quid faciebat obsecro? (ASD 1.2: 639)

So like a Roman did this Roman speak, that I heard nothing about the Death of Christ. Yet nevertheless this most ambitious proponent of the Ciceronian manner of speaking, seemed to the Ciceronians to have spoken marvelously, although about the subject, which he seemed neither to understand nor to care for he spoke almost nothing, and was not saying anything appropriate, nor had he stirred any feelings. Only this portion of praise was he winning, that he had spoken like a Roman and recalled something of Cicero. This would be approved as a proof of quality and for those who were clever, if by a boy to other boys in school such a speech had been given. But what did it accomplish on such a day, to such an audience, to such a subject, I ask you?

Erasmus was following the humanist curriculum, but trying to teach for acquisition and for active use of the language. He saw the Ciceronian humanists, with their insistence on style over content, as opposed to the kind of communication-focused teaching he was promoting. There was a segment of Renaissance teaching theory which argued that learning should be hard; discipline was most important. Skills were

73 Rummel 1990: 123-137. Latin text from ASD 1971: 606-710.

Oakes 49 broken down and practiced piece by piece, built on one another. If students learned all the morphology by heart, they would be able to read. Every step of the process had to be perfectly comprehended and

“intuitive leaps of learning were distrusted” (Grendler 1989: 409). Intuitive leaps of learning are another way of describing the process of acquisition. Part of the natural order hypothesis is the acknowledgement that students will proceed through learning at their own pace, sometimes remaining in place and then making great strides quickly. More than that, acquisition is by definition a subconscious and thus intuitive process.

Following the Renaissance, the vernacular languages became literary languages in their own right and the fixedness of Latin was no longer a unique attribute. As Latin dropped from communicative use, the renaissance ideal of building language piece by piece according to logical rules, as had been practiced in the Renaissance, became the main focus of particular philological efforts and the Grammar-

Translation methodology dominated. Now standards for language learning are being rewritten again, but there are still lessons to be learned from this sixteen-hundred-year-long history of communicative

Latin language teaching.

Oakes 50

CHAPTER THREE: Historically Backed Comprehensible Input Methods in the Modern Classroom

A Focus on Sound

Through the sixteenth century, there is evidence in the history of Latin pedagogy that successful teachers were using practices we now ascribe to strategies associated with Comprehensible Input. Many of their practices were consistent with Krashen’s hypotheses and can be shown to support the goal of

Living Latin pedagogy. These practices have multiple applications in a modern classroom.

A substantial difference between modern methods and those of the historical periods discussed is the focus on sound, not simply on using oral Latin or understanding individual words, but on understanding the pronounced sounds and syllables that make up a Latin word and how those sounds relate to letters. In antiquity, Greek speakers had to learn a new alphabet. The first part of Alcuin’s grammar focused on the letter. Primers throughout the Middle Ages required students to be able to take dictation and to divide the sounds of words into syllables (Grendler 1989: 142-161). These skills, too, often are treated dismissively or not at all in most modern Latin textbooks. For example,

Wheelock’s Latin places its section on the topic, “The Alphabet and Pronunciation” in the last part of the

Introduction (Wheelock 2005: xxxiv-xliv). On the many modern message boards, blogs, and teaching groups that exist for teachers attempting Living Latin,74 I can find no specific mention of a focus on the individual sounds that make up words.

A modern Comprehensible Input classroom for early learners can introduce a focus on sounds

(not words) in many ways. Many teachers require “attention getters” to restore focus after group work or transition periods. They might start a pattern of claps that the students then continue, such as the old

“SHAVE-and-a-HAIR-cut” rhythm. A repetition of Latin sounds and syllables might be used instead: a teacher might begin reciting the alphabet and continue until the students join in. They could also start with bo ba bu bi as the Renaissance primers did (Black 2001: 36-44) to help students understand how

74 See Appendix A.

Oakes 51 the letters sound when paired with the different vowels. When teaching a new word, the teacher can spell it using the Latin pronunciations of each letter rather than writing it on the board. Excessive focus on sound, syllables, and words as building blocks risks introducing merely conscious learning, but it is important that students develop the ability to separate meaningful syllables from non-meaningful ones.

As in the medieval monasteries, students today can be surrounded with the sounds of Latin: a teacher can play short amusing YouTube videos in Latin as they enter the class, play Latin chant music during quiet work time, or play recordings of different expert voices reading ancient texts.

The practice of dictation was also used extensively before access to texts became commonplace.

This practice required students to be able to group syllables into words and write those syllables with the proper letters, a reverse of the practice used when reading aloud. Students can demonstrate through a correct transcription that they understand what they are listening to, and teachers can check comprehension without forcing production. If a student is not placing the word breaks properly between syllables, e.g., writes “intabul amscri bere” instead of “in tabulam scribere,” then it will be obvious that comprehension is lacking. Dictation is used for this purpose in some Comprehensible Input classrooms, but the practice could be expanded now that our historical analysis has shown how effective an aid it can be for clarifying understanding.

Attention to Bilingual Sense Units

Beyond merely understanding what syllables make up words, the ancient bilingual columnar texts from the Hermeneumata show how words are combined into sense units. Making use of sense units can be an important method of focusing on comprehension. While this presentation was added to some ancient texts, a much larger portion of the bilingual texts (e.g. the colloquia) were initially written with this presentation in mind. The writer used sense units that could also be expressed in approximately the same number of words in the first language, though the grammar used to express the meanings may have varied between the two languages. Teachers can write exercises of this kind in

Oakes 52 early-stage Comprehensible Input. Such exercises need to prioritize sense units, but preserve meaningful, relatable content. Conversational Latin is much more inclined to feature shorter sense units and can be more meaningful to students because it has an immediately obvious use.

The colloquia of Erasmus, Vives, and the Hermeneumata schoolbook illustrate the proper behavior for a schoolboy. Similar short, slightly humorous colloquia written with sense unit equivalence in mind could be developed for modern students, for example, to use on the first day of class to express classroom values; these colloquia could be repeated occasionally, or even every day. A collection of simple phrases might include:

Maturus ero. (I will be on-time.) Magistra “salve!” dicet. (The teacher will say “Hi!”) Sedebo (I will sit) in sedem meam. (in my place.) Stilum habebo. (I will have a pencil.) Finis hodierna (Today’s goal) scribetur. (will be written down.) Nos Discipuli (We as students) attendemus. (will listen carefully.) Auxilium rogabo. (I will ask for help.) Comites mei (My classmates) me audient. (will listen to me.) Audiam eos. (I will listen to them.) Simus aequissimi (Let us be very fair) inter nos (to each other.)75 Nemo dormiat. (May no one fall asleep.)

The meaning of these phrases could be taught either by acting them out or by quickly providing the translation. The important thing is not to focus on what each word means, but on what the words mean together. They are written in a way that prioritizes content over grammar, including passive and future constructions from whatever conjugation suits the meaning. More complicated colloquia, perhaps imitating the Hermeneumata schoolbook or Erasmus could be introduced so long as the idea of purposeful bilingual sense units is preserved.

75 The Latin Teacher Idea Exchange Facebook group (see Appendix A) supplied this equivalence.

Oakes 53

Early Introduction of Meter

The next step is to introduce students to a reader with short (i.e. regular), memorable meters.

Throughout the medieval periods and the Renaissance, Cato’s Disticha or the Psalter were used to introduce meter. These texts were useful for students because they were meaningful to them and the couplets made them easy to memorize. They were repetitive in structure as well as content, which helped make them comprehensible. They also served to introduce students to meter early in their Latin instruction, so that their first exposure to meter was not at the same time that they were reading the elevated content of most metrical texts. Many of the Disticha are still relevant to a student and might fit into the modern classroom in a number of ways. The important feature of the exercise is not memorization, but that the Disticha serve as input both for the student and their fellow students listening to the recitation:

1.2 Plus vigila semper nec somno deditus esto; Nam diuturna quies vitiis alimenta ministrat.

Always be more vigilant and do not surrender to sleep; For continual rest supplies nourishment to vices.

1.16 Multorum cum facta senex et dicta recenses, Fac tibi succurrant, iuvenis quae feceris ipse.

When, as an old man, you count the deeds and sayings of others, Make what things you did yourself as a youth support you.

1.32 Ignotum tibi tu noli praeponere notis; Cognita iudicio constant, incognita casu.

Do not prefer the unknown over the known; Known things depend upon judgement; unknown things, upon chance.76

Memorizing and understanding a series of these couplets, as opposed to short prose sentences or quotations from ancient authors, are made easier because many of them follow a similar format. The

76 Latin text from http://www.thelatinlibrary.com/cato.dis.html.

Oakes 54 pattern of 1.16 (a circumstantial clause in the first line, advice in the second line) appears in fifty of the couplets (Thomson and Perraud 1990: 52-53). Such formulaic content becomes tedious once mastered, but the repetition is extremely useful when students are getting used to a new author’s style or content.

The idea of short meter warm-ups is not new to Comprehensible Input,77 but the texts in use often contain selections from many authors and periods. Though we have seen that survey courses containing varied selections may be more useful at higher levels, as soon as students are able to read and pronounce Latin, they can appreciate meter and understand it through narrow, but interesting content. Providing early and rewarding experience with meter is part of pre-teaching/sheltering, in order to keep the affective filter low when more complicated meter is introduced. In Alcuin’s grammar, his schoolboys quote the Aeneid; this, too, is a kind of pre-teaching for the Aeneid text the schoolboys will read later.

Grading by Content rather than Grammar

Other historically-based methods to prepare students for the Aeneid include using prose summaries, such as those in the Hermeneumata for Homeric passages, or the paraphrase commentaries introduced in the Renaissance. These exercises resemble embedded readings, a practice of teaching the same content repeatedly via increasingly difficult grammar constructions (Patrick 2015: 120-22).

Embedded readings gradually increase the difficulty of the grammar; a more medieval approach is to instead grade the level to which students are expected to interact with the text.

Consider this sample progression a teacher might follow to provide a new and different level of exposure to the Aeneid every year. The first time a teacher introduces a portion of the Aeneid, students are asked to listen to the words and get used to the sound of the meter. They perhaps will find some vocabulary words familiar. Teachers can use an understandable prose paraphrase to provide the

77 See Appendix A, especially Laura Gibbs’ blog.

Oakes 55 meaning rather than expect the students to understand the text directly. One such quote that is easily relevant to a classroom environment on a day when the students need some extra encouragement is:

experti: revocate animos, maestumque timorem mittite: forsan et haec olim meminisse iuvabit. Virgil, Aeneid 1.202-3

[You] have been tested: revive your spirits and send away your sadness and fear: perhaps it will please you to remember even these things one day.

A general idea of the passage in prose that the teacher would expect the students to understand:

Aeneas dixit, “vos parati estis. Nolite timere! postmodum melius tempus erit!” (‘Aeneas said, “you are prepared. Don’t be afraid! A better time will come!”‘).

The second time students meet the Aeneid may be in a simple prose story form, similar to those the Hermeneumata contain for Homeric passages, the story of Aeneas and Dido (Aeneid 4) lends itself to such a summary. Students might be asked to write or otherwise share the story a summary of the story to demonstrate comprehension. Rather than expecting the students to use perfectly classical Latin, the teacher should encourage students to attempt the exercise using their own learner language which would allow for errors and grammatical structures appropriate to the ordo naturalis. For those students who are concerned with correctness, the teacher can aid them in using their monitor as an editor. As the students share their summaries, they provide more and different input on the same content to their fellow students.

In the third year, students meet syntactically-chunked selections of the Aeneid containing descriptions or imagery, such as a portion of Aeneas’ vision of the death of Priam (2.506-517), or the scenes on Aeneas’s shield (8.630-70), and are invited to draw these narrative fragments. They subsequently could be given selections regarding characterization, such as Aeneas’ meeting with Dido in the underworld (6.437-473). Eventually they will be able to understand the entire narrative of the

Aeneid in its original form and react to it in more complex ways.

Oakes 56

Throughout the process the teacher provides paraphrase commentaries to pre-teach some content and vocabulary before presenting an unadapted text; afterwards, paraphrased commentaries may serve as a means of checking comprehension. Following medieval tradition, the teacher provides

Latin synonyms for difficult vocabulary or else quick English translations, as many as are necessary to aid students in comprehending the text. These difficulty of successive exposures increases slowly, along with the students’ facility with the content or faculty for higher level thinking. The challenge for the teacher is to find or create content that can be sheltered and is within range of the students’ i+1. Still, throughout the student’s Aeneid experience, the focus is on learning about the Aeneid as a story or work of literature, not on improving translation skills.

Accepting Learner Language and Non-Classical Vocabulary

Augustine and Erasmus both commented on the need to let curiosity drive student learning rather than punishment (e.g., a poor grade, in our modern context). In the later Middle Ages, many non- classical topics were discussed in Latin and new coinages of words were necessary. When the humanists limited themselves to classical vocabulary, they either were confined to classical topics or paraphrases that were difficult to comprehend. Allowing coinages respects a student’s curiosity and provides opportunities for self-expression for. Allowing the use of a learner language, on a lower register like the ordo naturalis, respects each student’s level of acquired language. Lack of explicit correction avoids attempting to force acquisition of a feature which is not next in the language’s natural order. Following the Jesuits, a teacher can encourage creative writing and transformation projects or competitions which teach the students something about both the writing process and what they are imitating. Students could turn passages from Ovid into plays and perform them in class or create short persuasive speeches based on one of Horace’s satires. Accepting learner language and non-classical vocabulary invites students to engage at their own level as they communicatively learn historical and cultural content.

Oakes 57

Lingua Latina per se Illustrata (Ørberg’s 2003 revision of his original Natural Method text) includes chapters that are loosely based upon Roman culture. (For example, Book 1, Chapter XII, is based upon the Roman calendar, and teaches the months and years). A fault of this text is that, while it provides repetitive and mostly comprehensible input, there is still a focus on form over content. The important information students are meant to acquire from the text is the new grammar that is introduced and that is what the majority of the exercises test. The exercises feature sentences drawn from the chapter; they ask students to supply missing endings or words. There is also a series of comprehension questions which ask students to demonstrate that they have understood the content of the text, e.g. Quot sunt menses anni? (‘How many are the months of the year?’) (Ørberg 2003: 102). The text does not ask students to do anything with the content they are learning in order to make it meaningful for themselves.

Krashen’s hypothesis about Comprehensible Input is based upon the idea that students have to find content meaningful in order for them to comprehend it as language input. Alcuin’s classroom might have tried to find a moral lesson or the underlying universal order from their study of the Roman calendar. The humanist classroom might have been concerned with Julius’ and Augustus’ innovations and how that reflected their power. A current classroom might incorporate a sundial building project to help understand the Romans’ knowledge of basic astronomical science, or to determine how the structure of the day affected social behavior. Such lessons would require the incorporation of new coinages and vocabulary unknown to the Romans. It takes a great deal of work on a teacher’s part to design these kinds of lessons for the particular interests of their students, but the focus on learning meaningful content through the Latin language was one primary way that Latin stayed relevant through the different historical periods. The following sample lesson, in which students create a personalized

Roman history project, incorporates some of the Comprehensible Input practices discussed above.

A Sample Lesson

Oakes 58

Pre-class: At the door, the teacher greets the students in Latin and asks how they are doing that day. If a student needs prompting, the teacher models this by offering three response options, using hand gestures: thumbs up with bene habeo (‘It’s going well’); waggling hand palm down with varie habeo (‘I’m up and down.’); and thumbs down with male mihi est (‘It’s not going well for me.’). Students can respond with either the gesture or the phrase. As the students sit down, a Latin song from the

Carmina Burana is playing quietly.78 As the students settle in, they read and write down the plan for the day, including the lesson objectives and the next homework assignment, both of which are written in bilingual sense units, Latin and English, on the whiteboard. There are also a number of useful Latin phrases, to be used during the class meeting, written on the board

Start of Class: The teacher greets the class as a whole in Latin and explains what the plan is for that day. The teacher both uses the Latin phrases already on the board for the objective and homework, and offers paraphrases or alternate glosses of the same information.

Review Lesson: The teacher introduces the class activity using Latin. The teacher has prepared cardstock pages for events in Roman history. One side of each page has a visual cue for the event, the other has a hexameter couplet and short prose paragraph describing the event. Students are each given one card and asked to arrange themselves in a “human timeline” around the room. The students are encouraged to provide information from the text on their card to each other and ask simple questions in

Latin until everyone is arranged in the correct order. This interactive activity is designed to get the students out of their seats and speaking to one another in Latin. The teacher may help the class get started by modeling the kind of questions to ask. Once the students are in the correct order, each

78 Carl Orff’s classical composition performed by UC Davis Symphony Orchestra, the University Chorus and Alumni Chorus, and the Pacific Boychoir at the Mondavi Center at UC Davis is available on YouTube. https://www.youtube.com/watch?v=QEllLECo4OM The selection “amor volat undique,” which begins at 43:31 is has clear Latin pronunciation.

Oakes 59 student names his or her event and describes it briefly in Latin, either by reading a portion of the provided text, or sharing a Latin summary.

Start of New Content: The students are still standing in a circle or half circle at this point. The teacher then pulls out a new card with a visual of an additional historical event, without a description.

The teacher fits herself into the timeline, provides the date of the event, and describes the place of this event in relation to the other events, e.g. this happened before Augustus’ reign but after the Punic

Wars; this happened during the second triumvirate, but between the death of Antony and the death of

Cleopatra. The teacher can also add humor by using the students’ names instead of the historical cards they are holding, e.g. this happened before Penelope, but after Charlie; Rebecca was partly responsible for making it happen; Valerie was one result of this event. Because the students are holding the cards with the events visible, it will still be clear what events are being referenced.

Group Work: The teacher collects the cards. Students return to their desks and are divided into groups. Each group is given a copy of the card for the new event and the same piece of classical Latin prose text describing the event. The teacher reads the Latin text to the whole class and provides a paraphrase commentary. The students are allowed to take notes from the teacher’s commentary and annotate their passages. Students are then asked to discuss the event and decide as a group what about it was most important. They work together to create their own prose summary or hexameter couplet to describe it. After the groups are finished, each group shares the summary or couplet with the rest of the class. The teacher collects the students’ work. After class, the teacher can edit the students’ contributions to write what will be on the back of the final event card to be used in future classes.

End of Class: The students write in their language journals a few sentences about what they learned that day in class related either to the history content or the language content, how well they think that they fulfilled the day’s lesson objectives, and any questions they still have. A few students are offered the opportunity to share what they journaled. The teacher dismisses the class. She then stands

Oakes 60 at the door and asks her students to tell her in English or Latin the most important thing they learned today.

Oakes 61

CONCLUSION

As part of the Living Latin movement, the current draft Standards for Classical Language

Learning (ACL 2016) view the purpose of teaching classical languages in ways similar to the antique and medieval language teachers. The Standards forefront acquisition over learning by setting proficiency

(displays of language ability) rather than performance (expression of guided or practiced language) levels to measure student growth. Performance can be taught and is tied closely to the curriculum a teacher employs. Grammar-Translation courses often test students on passages that have already been read in class with the aid of commentaries, and with the teacher sheltering the instruction. A student can perform well on this type of assessment, but be unable to read a portion of the same text which was not covered in class. Proficiency on the other hand, is independent of students’ ages or the curriculum they have followed. Reading proficiency is demonstrated only by the students’ ability to read a text at sight.

The Standards represent Latin as a tool to open students to new ideas and content. The

Standards include five goal areas: Communication, Cultures, Connections, Comparisons and

Communities. While all the goals are interconnected, the Standards explicitly state: “The communication goal guides students and teachers in the process of acquiring language skills, which [skill] then enables them to explore the rich areas available to them in the remaining goal areas” (ACL 2016: 3).

Within the communication goal, the Standards describe three different subsets: Interpretive

(what does the producer of the language want you to understand?), Interpersonal (“exchanging information, reactions, feelings and opinions by negotiating and clarifying meaning”), and Presentational

(communication used for a specific purpose and audience) (ACL 2016: 7). Using only Grammar-

Translation methodologies, it is very difficult to meet any of these communication goals, let alone the four other goals to which the language is meant to give access. Complementing current Comprehensible

Input methodology with techniques borrowed from history promotes communication.

Oakes 62

In the first area, interpretive communication, the ability to read and understand (a traditional goal) is married to the ability to listen and understand. In antiquity, because of the lack of texts, taking dictation was a regular practice as students learned Latin. In the medieval monasteries, students were surrounded by oral Latin. Both teachers and students read aloud and were expected to understand the reading without decoding. Being able to understand the text read by the teacher was the first stage of learning. As other students presented their exercises, input at every level was available to the students and could be understood and perhaps commented on by the students. After reciting a colloquium or passage, summarizing what was just heard was a standard exercise. Such exercises have been shown throughout this thesis to be adaptable for a modern classroom. The interpretive communication goal is the most amenable to a Grammar-Translation classroom, but can only be met fully by a classroom that practices active oral-aural use of the language, as well as reading.

The interpersonal communication goal forefronts the idea that information is shared “by negotiating and clarifying meaning” (ACL 2016: 7), which is the same process that enables i+1. More and varied input is provided until comprehension is achieved. Teachers can model this clarification of meaning by constantly evaluating their students’ level of comprehension. The kind of classroom described in Alcuin’s grammar and the paraphrase commentaries from the Renaissance show methods that advance this kind of communication. Erasmus’ argument that students must like their teacher supports this goal as well. A positive student-teacher relationship facilitates communication in the moment when a student is ready to speak and wants to speak, as Augustine described in the account of his early language learning in Confessions 1.14. Negotiation of meaning can also be practiced between students who work together and present their ideas to each other.

The presentational goal respects the place of the monitor (learned grammar) in the classroom.

When preparing for presentations, students have time to edit and be concerned for correct language use, but lessons from the Renaissance show how easily this concern can lead students and teachers

Oakes 63 astray. The Standards describe presenting for a specific purpose and a specific audience. In their carefully structured Latin instruction sequence, the Jesuits had students consistently produce and perform creative works in different genres and styles. On the other hand, Erasmus described in his

Ciceronian how a strict focus on one style and one kind of correctness which ignores the actual goal of the presentation results in no actual communication taking place. When the focus is on students as they explore new content, there is a clear place for meaningful presentational communication in the modern classroom.

Above all else, the message from the new Standards, from Krashen, and from this study of historical Living Latin pedagogy, is that content must always be the goal. Ancient Greeks trying to learn

Latin usually wanted to be able to participate in a society in which Latin was the lingua franca. The language input they received was focused on providing them with the practical skills necessary to meet that goal. In the medieval period, after basic literacy was taught, the texts were graded not by the difficulty of grammar, but the difficulty of the content. Their goal was to educate moral and philosophical teachers, with the secondary goal of training for clerical positions. In the Renaissance, the goal was to provide young men with exemplars from the ancient world that they could follow to become leaders in civic life. The humanists went astray when they attempted to advance students’ acquisition faster than comprehension, and when their strict need for correctness engaged the monitor.

This impetus for correctness is part of what continues to hold people back from accepting

Comprehensible Input into their classrooms (Owens 2016: 515). Like the humanists who corrected medieval orthography and grammar according to the classical norms, modern scholars too often are on the prowl for errors. But Donatus and other early grammarians were also on the lookout for errors, and

Erasmus presents a Latin almost perfectly classical in its syntax, yet still able to be used in daily life.

When Latin was spoken widely, learners used whatever level of the language they had acquired to communicate in and out of the classroom. But in the modern classroom, problems arise when the

Oakes 64 learner-language is not accepted, but punished. When creative uses of the language and the creation of new vocabulary necessary for the topic are not allowed, communication may be impeded. Cicero left us no conversation about mutant-superheroes, but the need for a few coinages and a more conversational style when students compare and contrast their own heroic culture with that of the Romans does not mean the activity should have no place in the classroom. Latin can retain its basic grammar and syntax and still be used to communicate contemporary content. Barbarisms can be avoided through continuous use of good Classical Latin input, but must be accepted as students learn to explore the language. The goal is to understand classical Latin and use the language as a tool to enable that understanding. While

English speakers typically do not use highly rhetorical speech informally, they are still able to comprehend it when necessary. The same can be true of Latin; students will be able to read and understand texts which are far above their own level of production.

Another criticism of Comprehensible Input is also familiar from study of the Renaissance, namely the idea that Latin and indeed all learning ought to be difficult.79 Cheryl Lowe concluded her

2012 article against the Natural Method with the comment, “The natural method is not only contrary to common sense but to the whole premise of formal education, which is to teach by systematic, logical instruction.” She is not content with a classroom full of relaxed, playful students (she criticizes the songs employed by her son’s German teacher), but concerned more with speeding students through learning.

Indeed, the speed at which it is possible to acquire language is a valid criticism. Medieval students devoted themselves to Latin. In the Renaissance, Latin was the only subject studied for six hours a day from ages ten to thirteen, in Jesuit schools. A further two years were devoted to readings in the humanities (primarily history and moral teachings) and a final year to rhetoric (eloquence) (Grendler

1989: 377-78). In a modern Comprehensible Input classroom, even an excellent teacher will find it

“impossible to prepare all kinds of students to read classical literature in 3-4 years” (Patrick 2015: 132).

79 See pp. 49-50.

Oakes 65

Taking students through the AP curriculum in senior year using entirely Comprehensible Input methods can also remain out of reach even for the most skilled Comprehensible Input teacher (Patrick 2015:

117).80 Teachers might set their sights instead on having students achieve mastery, meet the Standards for Classical Language Learning, develop a love of the language, and wish to continue their study of

Latin.

The final question many teachers new to Comprehensible Input struggle with is the place of grammar. This is a question for more research. A general piece of advice from the ancients, however, is to focus only on the grammar that is necessary for understanding. Dositheus made the ablative section of his grammar bilingual, but left the cases that Latin has in common with Greek untranslated. Through the humanist period, grammars continued to treat only the basics of morphology and syntax.

Speculative grammars were trialed with adult learners in antiquity, but mostly unsuccessful in the secondary schools of the later Middle Ages. A focus on forms is still possible in Comprehensible Input classrooms when these are the forms necessary for communication, and methods for that focus are constantly being developed (Ellis 2006:102).

Mary Beard, the well-known classicist, who writes a regular column for the Times Literary

Supplement, once asked “Is Latin too Hard?” (2006). In that column, she dismissed the modern language focus on learning conversational skills, but acknowledged the intellectual difficulty of learning to read classical texts. She writes, “Asking a school student to read Tacitus is a bit like asking an English learner to go off and read Finnegan’s Wake. But it is what makes the whole enterprise intellectually worthwhile.” This very idea, that what we are asking of students is too difficult, is what inspires some

80 Patrick allows that some learners might excel and be capable of taking on the AP curriculum, but remarks that the AP course at his school is viewed as separate from the rest of the Comprehensible Input curriculum (2015: 117).

Oakes 66 teachers to embrace Comprehensible Input teaching.81 Establishing achievable goals is central to all good teaching. Mary Beard comments in her article that if learning about classical civilization or formal grammar is the goal, there are more efficient ways to do so than a Latin course.

There is something unique about reading a classical author: the humanists were certainly right in that regard: but the idea that students cannot enjoy the journey, that their own pace of learning cannot be accommodated, and that their own interests cannot motivate them, these are lessons from the humanists that we should not have embraced. That Classical Latin has continued to be taught as a second language for over 2000 years demonstrates that teachers have been doing something right over the centuries, and that the literature to which that instruction offers access remains valued. This thesis has attempted to identify those practices from the history of Latin pedagogy that are now ripe for imitation, and also to advance the case for wider implementation of Comprehensible Input practices in

Latin instruction.

81 See http://todallycomprehensiblelatin.blogspot.com/p/views-on-classical-literature.html (one teachers explanation of his opposition to the “mad rush” towards classical literature) or https://eidolon.pub/teaching-latin- to-humans-4e6b489b4e17 (another teacher’s request to honor “the Latin language and our students’ minds”).

Oakes 67

APPENDIX A: Resources for Living Latin Pedagogy

Bestiaria Latin Blog by Laura Gibbs: http://bestlatin.blogspot.com/

CAMWS, Teaching Classical Languages, “Special Section on Spoken Latin”: http://tcl.camws.org/sites/default/files/TCL%20Spring%202015.pdf

Divus Magister Craft: https://www.youtube.com/channel/UCTtKmPD0_Qo9Uy932ZGKFhA/feed

Indwelling Language with Justin Slocum Bailey: http://indwellinglanguage.com/limen-a-latin-teaching-portal/

Latin Teacher Toolbox: http://www.latinteachertoolbox.com/

Latin-Best Practices Yahoo Group: https://groups.yahoo.com/neo/groups/latin-bestpractices/info, https://latinbestpracticescir.wordpress.com/home/

Paidaia: http://www.paideiainstitute.org/living_latin

SALVI: http://latin.org/wordpress/

Schola Latina: http://scholalatina.it/en/

Teaching Latin for Acquisition Facebook Group: https://www.facebook.com/groups/AcquireLatin/

Todally Comprehensible Latin: http://todallycomprehensiblelatin.blogspot.com/

Oakes 68

APPENDIX B: Relevant Latin Textbooks by Method

Grammar-Translation: Wheelock’s Latin, Frederic M. Wheelock (1956) Inductive Reading Method: The Cambridge Latin Course, Cambridge University Press (1970) Latin via Ovid, Jacob E. Nyenhuis and Norma Goldman (1977) Ecce Romani: a Latin Reading Course, Scottish Classics Group (1982) Oxford Latin Course, Maurice Balme and James Morwood (1996) Direct Method: Linguaphone. The Direct Method Applied to Latin. A Handbook for Teachers. W.H.D. Rouse (1930) Natural Method: Lingua Latina, Hans Ørberg (1955) Lingua Latina per se Illustrata, Hans Ørberg (2003)

Oakes 69

BIBLIOGRAPHY

ACL 1924. The Classical Investigation, Part One: General Report. Princeton.

ACL 2016. Standards for Classical Language Learning, draft accessed: https://aclclassics.org/uploads/assets/files/Standards/Standards_for_Classical_Language_Learni ng_2016.pdf 18 April 2017.

Adams, J.N. 2003. Bilingualism and the Latin Language. Cambridge: Cambridge University Press.

ASD 1971. Opera Omnia Desiderii Erasmi Roterodami, Ordinis Primi, Tomus Secundus. Amsterdam: North-Holland Publishing Company.

Beard, M. 2006. “Is Latin too hard?”, The Times Literary Supplement, accessed: http://timesonline.typepad.com/dons_life/2006/06/is_latin_too_ha.html 30 Mar 2017.

Black, R. 2001. Humanism and Education in Medieval and Renaissance Italy: Tradition and Innovation in Latin Schools from the Twelfth to the Fifteenth Century. Cambridge: Cambridge University Press.

Brown, P. 1967. Augustine of Hippo. Berkeley: University of California Press.

Buehl, D. 2014. Classroom Strategies for Interactive Learning (4th ed.). Newark, DE: International Reading Association.

Campbell, F. 1968. “Latin and the elite tradition in education,” The British Journal of Sociology, 1968: 308-25.

Carlon, J. M. 2013. “The Implications of SLA Research for Latin Pedagogy: Modernizing Latin Instruction and Securing its Place in Curricula”, Teaching Classical Languages, 2013: 106-22.

Carr, W.L. 1928. “Shall We Teach Our Pupils to Read Latin?” Classical Journal 23: 500-10.

Carter, D. 2011. “Hans Oerberg and his contribution to Latin pedagogy,” The Journal of Classics Teaching, N. 22: 21-22.

Clackson, J. and Horrocks, G. eds. 2007. The Blackwell History of the Latin Language. Malden, MA: Blackwell Publishing.

Coffee, N. 2012. “Active Latin: Quo Tendimus?” The Classical World, 105: 255-69.

Oakes 70

Copeland, R. and Sluiter, I. 2009. Medieval Grammar and Rhetoric: Language Arts and Literary Theory, AD 300-1475. Oxford: University of Oxford Press.

Courchene, R.J.; Glidden, J.I.; St. John, J.; Therien, C. eds. 1992. Comprehension Based Second Language Teaching. Canada: University of Ottawa Press.

Cracas, L. 1970. “What is ‘Reading Latin as Latin’?” The Classical World, 64: 81-86.

Davis, S. 1991. Latin in American Schools: Teaching the Ancient World. American Philological Association. Atlanta: Scholars Press.

Demetriou, C. 2014. “Aelius Donatus and his Commentary on Terence’s Comedies,” in Fontaine and Scafuro, The Oxford Handbook of Greek and Roman Comedy. Oxford. 782-99.

Demetriou, C. 2015. “Donatus’ Commentary: The Reception of Terence’s Performance,” in Turner and Torello-Hill, Terence between Late Antiquity and the Age of Printing, Leiden and Boston: 182-99

Dickey, E. 2016. Learning Latin the Ancient Way: Latin Textbooks form the Ancient World. Cambridge: Cambridge University Press.

Dickey, E. (ed.) 2012. The Colloquia of The Hermeneumata Pseudodositheana, Vols. 1-2. Cambridge: Cambridge University Press.

Ellis, R. 2006. “Current Issues in the Teaching of Grammar: An SLA Perspective,” TESOL Quarterly, 40: 83- 107.

Farrell, J. 2001. Latin Language and Latin Culture: from Ancient to Modern Times. Cambridge: Cambridge University Press.

Grendler, P. F. 1989. Schooling in Renaissance Italy: Literacy and Learning, 1300-1600. Baltimore: The Johns Hopkins University Press.

Hunt, T. 1991. Teaching and Learning Latin in 13th-Century England, Vols. I-III. Cambridge: D.S. Brewer.

JACT and ARLT 2011. “Communicative approaches to teaching the classical languages,” The Journal of Classics Teaching, N. 22.

Jedan, D. and Troyanovich, J. 1974. “Pestalozzi and Oral Latin,” The Classical Outlook, 52.4: 37-38.

Oakes 71

Krashen, S. D. and Terrell, T. D. 1983. The Natural Approach: Language Acquisition in the Classroom. The Alemany Press/Pergamon Press.

Krashen, S. D. 1981. Second Language Acquisition and Second Language Learning. Oxford: Pergamon Press.

Krashen, S. D. 1985a. Inquiries & Insights: Second Language Teaching Immersion & Bilingual Education Literacy. Hayward, CA: Alemany Press.

Krashen, S. D. 1985b. The Input Hypothesis: Issues and Implications. London: Longman Group Limited.

LaFleur, R. ed. 1987. The Teaching of Latin in American Schools: A Profession in Crisis. American Classical League and the American Philological Association. Scholars Press.

Law, V. 1989. “Learning to Read with the Oculi Mentis: Virgilius Maro Grammaticus,” Literature and Theology, 3: 159-72.

Leonhardt, J. 2013. Latin: Story of a World Language, trans. Kenneth Kronenberg. Cambridge: Belknap Press.

Lindzey, G. 2015. “The Biduum Experience: Speaking Latin to Learn Latin,” Teaching Classical Languages, 2015: 72-107.

Lowe, C. 2012. “The Natural Method is Not Natural,” The Classical Teacher, 2012, accessed: https://www.memoriapress.com/articles/natural-method-not-natural/ 30 Mar 2017.

Macdonald, S. 2011 “Krashen & Second Language Acquisition (SLA) theory – a re-evaluation of how to teach classical languages,” The Journal of Classics Teaching, N. 22: 3-5.

Mahoney, A. 2004. “The Forms You Really Need to Know,” The Classical Outlook¸ 81: 101-05.

Masciantonio, R. 1988. “Stephen Krashen and the Classical Languages,” The Classical Journal, 84: 53-56.

Menzer, M. J. 2004. “Aelfric’s English Grammar,” The Journal of English and Germanic Philology, 103: 106-24.

Migne, J.P. ed. 1844-65. Patrologia latina, cursus completus. 221 vols. Paris. Accessed: http://patristica.net/latina/

Mitchell, B.W. 1935. “Reading Latin and Writing Latin Once More,” The Classical Weekly, 28: 193-96.

Oakes 72

Morris, S. 1966. Vita Novae: New Techniques in Latin Teaching. Great Britain: Hulton Educational Publications.

Murphy, J. J. 1980. “The Teaching of Latin as a Second Language in the 12th Century,” Historiographia Linguistica, 7: 159-75.

NCSSFL 2003. “The Role of Latin In American Education: A Position Paper from the National Council of State Supervisors of Foreign Languages,” The Classical Outlook, 80: 147-49.

Ostler, N. 2007. Ad Infinitum: A Biography of Latin. New York: Walker & Company.

Owens, P. M. 2016. “Barbarianisms at the Gate: An Analysis of Some Perils in Active Latin Pedagogy,” Classical World, 109: 507-23.

Patrick, R. 2015. “Making Sense of Comprehensible Input in the Latin Classroom,” Teaching Classical Languages, 2015: 108-36.

Reynolds, S. 1996. Medieval Reading: Grammar, Rhetoric and the Classical Text. Cambridge: Cambridge University Press.

Richards, J. and Rogers, T. 2014. Approaches and Methods in Language Teaching, 3rd edition, Cambridge: Cambridge University Press.

Robey, D. 1980. “Humanism and Education in the Early Quattrocento: the De ingenuis moribus of P.P. Vergerio,” Bibliotheque d’Humanisme et Renaissance, 42: 27-58.

Ruebel, J.S. 1996. “The Ablative as Adverb: Practical Linguistics and Practical Pedagogy,” The Classical Journal, 92: 57-63.

Rummel, E. Ed. 1990. The Erasmus Reader. Toronto: University of Toronto Press.

Stray, C. 2011. “Success and Failure: W.H.D. Rouse and direct-method Classics teaching in Edwardian England,” in The Journal of Classics Teaching, N. 22: 5-7.

Thomas, M. 1995. “Medieval and Modern Views of Universal Grammar and the Nature of Second Language Learning,” The Modern Language Journal, 79: 345-55.

Thompson, C. R. ed. 1978. Collected Works of Erasmus: Vol. 24, De Copia De Ratione Studii. Toronto: University of Toronto Press.

Oakes 73

Thompson, C. R. Ed. 1997. Collected Works of Erasmus: Vol. 39 and 40, Colloquies. Toronto: University of Toronto Press.

Thomson, I. and Perraud, L. 1990. Ten Latin Schooltexts of the Later Middle Ages: Translated Selections. Wales: Edwin Mellen Press Ltd.

Too, Y. L. and Livingstone, N. eds. 1998 Pedagoy and Power: Rhetorics of Classical Learning. Cambridge: Cambridge University Press.

VanPatten, B. 1996. Input Processing and Grammar Instruction. Norwood, NJ: Ablex Publishing Corporation.

VanPatten, B. 2003. From Input to Output: A Teacher’s Guide to Second Language Acquisition. Boston, MA: The McGraw-Hill Companies, Inc.

Waquet, F. 2001. Latin or the Empire of a Sign: from the Sixteenth to the Twentieth Centuries, trans. John Howe. London and New York: Verso.

Wills, J. 1998. “Speaking Latin in Schools and Colleges,” The Classical World, 92: 27-34.

Wheelock, F.M. and LaFluer, R. rev. 2005. Wheelock’s Latin (Sixth Edition Revised), New York: HarperCollins Publishers.

Oakes 74